Full text of "VIVARIUM"

See other formats

Vivarium

Volume 45

2007

Reprinted with the permission of the original publisher

Periodicals Service Company
Germantown, NY
2009

Printed on acid-free paper.

This reprint was reproduced from the
best original edition copy available.

NOTE TO THE REPRINT EDITION:

In some cases full page advertisements which do not add to
the scholarly value of this volume have been omitted.

As a result, some reprinted volumes may have irregular pagination.

VIVARIUM

An InternationalJournalfor the Philosophy and Intellectual Life of the Middle
Ages and Renaissance

Aims & Scope

Vivarium is an international journal dedicated to the history of philosophy and the history
of ideas from the early Middle Ages to the early-modern period. It takes a particular interest
in the profane side of philosohy and its relationship with other areas of thought and learn¬
ing from these periods. It is widely recognized as an unrivalled resource for the history of
logic, semantics and metaphysics. It publishes philosophical analyses as well as historical
studies of ideas, texts and the institutional context of medieval and early-modern thought
and learning. It also welcomes editions of texts. It publishes annually a special issue devoted
to a particular theme or philosopher.

Editors

L.W. Nauta (Groningen), L.M. de Rijk (Leiden), H.A.G. Braakhuis (Nijmegen), C.H.
Kneepkens (Groningen), W.J. Courtenay (Madison), E.P. Bos (Leiden) and D. Perler
(Berlin).

Advisory Committee

T. Gregory (Rome), A. Zimmermann (Cologne), J.E. Murdoch (Cambridge, MA).

Notes for Contributors

Contributions to the journal should be sent as an e-mail attachment and paper version to
Lodi Nauta, Faculty of Philosophy, University of Groningen, Oude Boteringestraat 52,
9712 GL Groningen, The Netherlands (l.w.nauta@rug.nl). A style sheet is available at
www.brill.nl/viv or can be obtained from the editor.

Contributions must be written in English (UK or US), French or German; they must be
grammatically and stylistically correct. Manuscripts must be clearly typewritten with num¬
bered pages, double spacing and wide margins throughout. Use footnotes, not endnotes.
Title and name of the author should appear on a separate title page in order to facilitate
anonymous peer-review. Each article should be accompanied by a 10-line abstract, which
should state the principal conclusions of the paper and 2-6 keywords, for indexing pur¬
poses; both abstract and keywords should be in English.

Authors of accepted contributions receive one set of proofs for proofreading. The pub¬
lisher reserves the right to charge authors for corrections made to the proofs other than
corrections of errors that occurred in the typesetting or conversion process.

Authors receive 1 complimentary copy of the issue in which their article appears (in the
case of multi-authored contributions, copies are sent to the first-named author) as well as a
PDF file of their contribution for private circulation.

Vivarium (print ISSN 0042-7543, online ISSN 1568-5349) is published 3 times a year
by Brill, Plantijnstraat 2, 2321 JC Leiden, The Netherlands, tel +31 (0)71 5353500,
fax+31 (0)71 5317532.

VIVARIUM

An International Journal
for the Philosophy and Intellectual Life
of the Middle Ages and Renaissance

VOLUME XLV (2007)

BRILL

LEIDEN • BOSTON

BRILL

LEIDEN • BOSTON

© 2007 by Koninklijke Brill NV, Leiden, The Netherlands
Koninklijke Brill NV incorporates the imprints BRILL, Hotei Publishing,
IDC Publishers, Martinus Nijhoff Publishers and VSP.

All rights reserved. No part of this publication may be reproduced, translated, stored
in a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission of the publisher.

Authorization to photocopy items for internal or personal
use is granted by the publisher provided that
the appropriate fees are paid directly to Copyright
Clearance Center, 222 Rosewood Drive, Suite 910,

Danvers, MA 01923, USA. Fees are subject to change.

Printed in the Netherlands (on acid-free paper).

BRILL

Vivarium 45 (2007) 393-394

R1 UM

www.brill.nl/viv

Contents Volume 45 (2007)

Toivo J. Holopainen Anselms Argumentum and the Early

Medieval Theory of Argument . 1

Michael W. T kacz Albert the Great and the Revival of Aristode’s

Zoological Research Program . 30

Brian Francis Conolly Averroes, Thomas Aquinas and Giles of

Rome on Flow This Man Understands ... 69

Ernesto Perini-Santos La structure de Facte intellectif dans les

theories ockhamiennes du concept . 93

Luc Deitz Francesco Patrizi da Cherso’s Criticism of

Aristotle’s Logic . 113

Review . 125

Books Received . 128

Special Issue:

The Many Roots of Medieval Logic:

The Aristotelian and the Non-Aristotelian Traditions
Guest Editor:

John Marenbon

John Marenbon Introduction . 131

I. Roots, Traditions and the Multiplicity of Medieval Logic

Sten Ebbesen The Traditions of Ancient Logic-cum-

Grammar in the Middle Ages—What’s

the Problem? . 136

II. Stoic Logic and Linguistics

Christopher J. Martin Denying Conditionals: Abaelard and the

Failure of Boethius’ Account of the

ITypothetical Syllogism . 153

Martin Lenz Are Thoughts and Sentences

Compositional? A Controversy between
Abelard and a Pupil of Alberic on the
Reconciliation of Ancient Theses on
Mind and Language . 169

DOI: 10.1163/156853407X236776

394

Contents Volume 45 (2007) / Vivarium 45 (2007) 393-394

Anne Grondeux Res Meaning a Thing Thought:

The Influence of the Ars donati . 189

III. Platonism in Logic and Semantics

Christophe Erismann The Logic of Being: Eriugena’s Dialectical

Ontology . 203

Irene Rosier-Catach Priscian on Divine Ideas and Mental

Conceptions: The Discussions in the
Glosulae in Priscianum , the Notae
Dunelmenses , William of Champeaux
and Abelard . 219

Stefania Bonfiglioli Symbolism and Linguistic Semantics.

and Costantino Marmo Some Questions (and Confusions) from

Late Antique Neoplatonism up to

Eriugena . 238

IV. Aristotle, Augustine and Stoicism

Mary Sirridge “Utrum idem sint dicere et intelligere sive

videre in mente”: Robert Kilwardby,

Quaestiones in librum primum

Sententiarum . 253

Claude Panaccio Mental Language and Tradition

Encounters in Medieval Philosophy:

Anselm, Albert and Ockham . 269

Laurent Cesalli Intentionality and Truth-Making:

Augustine’s Influence on Burley and
Wyclif’s Propositional Semantics . 283

V. Aristotelian Traditions in Medieval Logic

Luisa Valente Names That Can Be Said of Everything:

Porphyrian Tradition and ‘Transcendental’

Terms in Twelfth-Century Logic . 298

E. Jennifer Ashworth Metaphor and the Logicians from Aristotle

to Cajetan . 311

Christophe Grellard Scepticism, Demonstration and the

Infinite Regress Argument (Nicholas of

Autrecourt and John Buridan) . 328

Catarina Dutilh Novaes Theory of Supposition vs. Theory of

Fallacies in Ockham . 343

Egbert P. Bos Richard Billingham’s Speculum puerorum ,

Some Medieval Commentaries and
Aristotle . 360

Bibliography 374

BRILL

Vivarium 45 (2007) 1-29

VIVA

RIUM

www.brill.nl/viv

Anselm’s Argumentum and the Early
Medieval Theory of Argument

Toivo J. Holopainen

Department of Systematic Theology, University of Helsinki

Abstract

The article aims at elucidating the argumentation in Anselms Proslogion by relating
some aspects of it to the early medieval theory of argument. The focus of the analysis
is on the “single argument” ( unum argumentum ), the discovery of which Anselm
announces in the Preface to the Proslogion. Part 1 of the article offers a preliminary
description of the single argument by describing the reductio ad absurdum technique
based on the notion “that than which a greater cannot be thought”. Part 2 discusses
the ideas about arguments and argumentation that Boethius presents in Book One of
his In Ciceronis Topica. Part 3 draws attention to some early medieval sources (Abelard,
Lanfranc, Anselm) that are witness to the importance of the Boethian ideas in Anselms
time. Finally, Part 4 argues that Anselm looked at his single argument in the Boethian
framework and that the term “that than which a greater cannot be thought” should be
identified as his single argument.

Keywords

Anselm, ontological argument, the single argument, Boethius, topics, theory of argument

Introduction

Anselms Proslogion is undoubtedly one of the most renowned and most dis¬
puted works in the history of Christian philosophical thought. The treatise is
famous for two features: the so-called ontological argument for Gods exis¬
tence and the programme of faith seeking understanding (fides quaerens intel -
lectum ). These are also the issues which make the treatise controversial. The
conventional outlook says that Anselm meant his argument for Gods exis¬
tence as a philosophically valid demonstration and he aimed at making the
content of faith understandable to the universal human reason. On the other
hand, the Proslogion is a devotional exercise in which the person who speaks

DOI: 10.1163/156853407X183171

77 J. Holopainen / Vivarium 45 (2007) 1-29

endeavours to elevate his/her mind to the contemplation of God in prayer.
This fact has made many scholars suspicious of the conventional view, and it
has been challenged by fideistic and mystic interpretations of Anselms thought.
In spite of intensive research dedicated to Anselm’s Proslogion during recent
decades, the correct interpretation of the treatise is still very much a controver¬
sial matter. The main reason for this is the peculiar combination of argumen¬
tation and devotion that the Proslogion exhibits, which makes it difficult to
perceive clearly how the argumentation in it should be construed. 1

The present article aims to elucidate the argumentation in the Proslogion by
relating some aspects of it to what early medieval dialectic has to say about
arguments and argumentation. The focus of the analysis will be on the “single
argument” (unum argumentum ), the discovery of which Anselm announces in
the Preface to the Proslogion} Part 1 offers a preliminary discussion of the
single argument. Part 2 discusses the early medieval theory of argument on the

1) The essays in The Many-Faced Argument: Recent Studies on the Ontological Argument for the
Existence of God, eds. John Hick and Arthur C. McGill (New York, 1967), can still serve as a
useful introduction to the controversial aspects of the Proslogion. In spite of its other merits, the
new general introduction, The Cambridge Companion to Anselm , eds. Brian Davies and Brian
Leftow (Cambridge, 2004), is not particularly useful for those interested in the interpretational
controversies surrounding the Proslogion , because there is very little explicit discussion about
those controversies in it. The essays in the volume reflect the mainstream views in the current
Anglo-American philosophical discussion. See also Richard Campbell, From Belief to Understand¬
ing: A Study of Anselms Proslogion Argument on the Existence of God (Canberra, 1976), and Greg¬
ory Schufreider, Confessions of a Rational Mystic: Anselms Early Writings (West Lafayette, 1994).

2) Even though there is no survey of the discussions about the structure of Anselms argumen¬
tation in The Cambridge Companion to Anselm there are some interesting references to Anselms
single argument in it. In their “Introduction”, Davies and Leftow say that “Anselm had much
more to offer about God than a single argument for His existence.” (p. 1). G. R. Evans, in her
essay “Anselms Life, Works, and Immediate Influence” (5-31, at 12-3), relates the story of
Anselms discovery and ends as follows: “There is a case for saying that the ‘argument’ he believed
he had discovered was a notion or principle which could be applied in a sequence of argumenta¬
tion or set like a jewel in a passage of prayer.” Evans fails to indicate whether such a case has
actually been made. In his contribution “Anselm’s Perfect-Being Theology” (132-56, at 140-1),
Brian Leftow offers an analysis of the reductio argument which Anselm uses, e.g., in Proslogion 5
and comments as follows: “This is an example of the Proslogion s ‘one argument’—really one
form of argument.” (Leftow’s emphasis.) These three references point in different directions when
it comes to the identification of the unum argumentum'. is it an argument for God’s existence, or
a notion or principle that can be applied in a sequence of argumentation, or a single form of
argument? Nevertheless, the Companion is not as incoherent regarding this matter as it may
appear from the above extracts. First, the view that the single argument is an argument for God’s
existence should not be seen as the view of the editors but as a reflection of the general opinion.
Second, the view that the single argument is a notion or principle can agree on quite many
points with the view that the single argument is a form of argument. The main difference is what
you take the signification of the term argumentum to be.

77 J. Holopainen / Vivarium 45 (2007) 1-29

basis of Boethius’s In Ciceronis Topica , whereas Part 3 draws attention to some
texts related to its reception in Anselm’s time. Part 4 returns to the single argu¬
ment in the Proslogion and offers a more precise depiction of how Anselm
understood it. 3

1. The Single Argument in the Proslogion

In Proslogion 2, Anselm presents a piece of argumentation which is known as
the earliest formulation of the ontological argument for God’s existence. Even
though it would be a mistake to identify this piece of text as the single argu¬
ment mentioned in the Preface, it can well serve as a starting point for a dis¬
cussion about the argument, because it is the text in which Anselm first makes
use of his argument. There Anselm introduces the idea that God is “something
than which a greater cannot be thought” and presents an argument in two
stages to demonstrate that such a being exists. In the first stage, he argues that
this thing exists at least “in the understanding” (in intellectu ), because when
someone hears the expression, he understands it, and whatever is understood
is in the understanding. 4 In the second stage, he further argues that “that than
which a greater cannot be thought” exists also in reality:

But surely that than which a greater cannot be thought cannot be only in the understand¬
ing. For if it were only in the understanding, it could be thought to exist also in reality—
something which is greater [than existing only in the understanding]. Therefore, if that
than which a greater cannot be thought were only in the understanding, then that than
which a greater cannot be thought would be that than which a greater can be thought! But
surely this [conclusion] is impossible. Hence, without doubt, something than which a
greater cannot be thought exists both in the understanding and in reality. 5

Some features of this reasoning are noteworthy from the point of view
of understanding Anselm’s single argument. Let me point out four such
features:

3) For a shorter and different version of the story, see Toivo J. Holopainen, Dialectic and Theology
in the Eleventh Century (Leiden, 1996), 133-45.

4) Anselm, Proslogion 2, ed. F. S. Schmitt, Opera omnia (Edinburgh, 1946-1961), 1:101: “Ergo,
domine, qui das fidei intellectum, da mihi, ut quantum scis expedire intelligam, quia es sicut cre-
dimus, et hoc es quod credimus. Et quidem credimus te esse aliquid quo nihil maius cogitari
possit. An ergo non est aliqua talis natura, quia ‘dixit insipiens in corde suo: non est deus’? Sed
certe ipse idem insipiens, cum audit hoc ipsum quod dico: ‘aliquid quo maius nihil cogitari potest’,
intelligit quod audit; et quod intelligit in intellectu eius est, etiam si non intelligat illud esse...”

5) Ibid., 101-2: “Et certe id quo maius cogitari nequit, non potest esse in solo intellectu. Si enim
vel in solo intellectu est, potest cogitari esse et in re, quod maius est. Si ergo id quo maius cogitari

T. J. Holopainen / Vivarium 45 (2007) 1-29

1. The inference represents the type of reasoning known as indirect proof or reductio ad
absurdum. Anselm takes the negation of the desired conclusion as an assumption and
then deduces a contradiction. Because of the contradiction, one can conclude that the
assumption is false and its negation true.

2. The contradiction deduced is, basically, that that than which a greater cannot be
thought is not that than which a greater cannot be thought.

3. The contradiction derives from a comparison between that than which a greater
cannot be thought as having a certain attribute or predicate and that than which a
greater cannot be thought as not having that attribute or predicate. In this case, the
comparison is between that than which a greater cannot be thought “not being
only in the understanding” and “being only in the understanding”.

4. The comparison that Anselm makes is in terms of greatness. Anselm presumes
that “not being only in the understanding” is greater than “being only in the
understanding”.

Similar features appear also in the other passages of the Proslogion , where
Anselm uses the notion “that than which a greater cannot be thought” as a
basis of argumentation. In Ch. 3, 5 and 15, Anselm makes the reductio quite
explicit, whereas in Ch. 18 it is only implied. The conclusions deduced are
that God, or that than which a greater cannot be thought, “cannot be thought
not to exist” (Ch. 3), is “that which—as highest of all things, alone existing
through Himself—made all other things from nothing” (Ch. 5), is “some¬
thing greater than can be thought” (Ch. 15) and is absolutely one and
indivisible (Ch. 18). 6

non potest, est in solo intellects id ipsum quo maius cogitari non potest, est quo maius cogitari
potest. Sed certe hoc esse non potest. Existit ergo procul dubio aliquid quo maius cogitari non
valet, et in intellectu et in re.” Trans. Jasper Hopkins and Herbert Richardson in Complete Philo¬
sophical and Theological Treatises of Anselm of Canterbury (Minneapolis, 2000), 93-4.

6) Proslogion 3, ed. Schmitt 1:102-3: “Quod utique sic vere est, ut nec cogitari possit non esse.
Nam potest cogitari esse aliquid, quod non possit cogitari non esse; quod maius est quam quod
non esse cogitari potest. Quare si id quo maius nequit cogitari, potest cogitari non esse: id ipsum
quo maius cogitari nequit, non est id quo maius cogitari nequit; quod convenire non potest. Sic
ergo vere est aliquid quo maius cogitari non potest, ut nec cogitari possit non esse.” Ibid., 5, 104:
“Quid igitur es, domine deus, quo nil maius valet cogitari? Sed quid es nisi id quod summum
omnium solum existens per seipsum, omnia alia fecit de nihilo? Quidquid enim hoc non est,
minus est quam cogitari possit. Sed hoc de te cogitari non potest.” Ibid., 15, 112: “Ergo domine,
non solum es quo maius cogitari nequit, sed es quiddam maius quam cogitari possit. Quoniam
namque valet cogitari esse aliquid huiusmodi: si tu non es hoc ipsum, potest cogitari aliquid
maius te; quod fieri nequit.” Ibid., 18, 114: “Quomodo ergo, domine, es omnia haec? An sunt
partes tui, aut potius unumquodque horum est totum quod es? Nam quidquid partibus est
iunctum, non est omnino unum, sed quodam modo plura et diversum a seipso, et vel actu vel
intellectu dissolvi potest; quae aliena sunt a te quo nihil melius cogitari potest. Nullae igitur partes

77 J. Holopainen / Vivarium 45 (2007) 1-29

It is possible to perceive this argumentative pattern in the Proslogion , if one
knows to look for it, and it is readily believable that the pattern has something
to do with the single argument. To get a more precise understanding of that
argument, however, one must leave the actual text of the Proslogion and turn
to the material surrounding it: the Preface to the Proslogion , Anselms reply to
Gaunilos critique, and, Anselm’s first treatise, the Monologion. Since the
Preface to the Proslogion is the only place where Anselm explicitly speaks of the
single argument, it is there where we will have to look first.

Anselm starts in the Preface to the Proslogion by making reference to his first
treatise. Having noticed that the Monologion was composed “of a chain of
many arguments”, Anselm says he began to ask himself

whether perhaps a single argument could be found which would require no other [argu¬
ment] than itself for proving itself and which would suffice by itself to demonstrate (1) that
God truly [i.e., really] exists and (2) that He is the Supreme Good (needing no one else, yet
needed by all [else] in order to exist and to fare well) and whatever [else] we believe about
the Divine Substance. 7

Anselm continues by talking about his desperate search for the argument and
about his joy when he finally discovered it. Then there follows a sentence
about the work he wrote:

Supposing, then, that if what I rejoiced to have discovered were written down it would
please its readers, I wrote the following work on this, and on various other [subjects], in the
role of someone endeavoring to elevate his mind toward contemplating God and in the role
of someone seeking to understand what he believes. 8

sunt in te, domine, nec es plura, sed sic es unum quiddam et idem tibi ipsi, ut in nullo tibi ipsi
sis dissimilis; immo tu es ipsa unitas, nullo intellectu divisibilis.”

7) Proslogion , Prooemium, ed. Schmitt 1:93: “... considerans illud esse multorum concate-
natione contextum argumentorum, coepi mecum quasrere, si forte posset inveniri unum argum-
tum, quod nullo alio ad se probandum quam se solo indigeret, et solum ad astruendum quia
deus vere est, et quia est summum bonum nullo alio indigens, et quo omnia indigent ut sint et
ut bene sint, et quaecumque de divina credimus substantia, sufficeret.” Trans. Hopkins and
Richardson, 88 slightly modified; Hopkins and Richardson here translate argumentum as
“consideration”.

8) Ibid., 93-4: “Aestimans igitur quod me gaudebam invenisse, si scriptum esset, alicui legend
placiturum: de hoc ipso et de quibusdam aliis sub persona conantis erigere mentem suam ad
contemplandum deum et quaerentis intelligere quod credit, subditum scripsi opusculum.” Trans.
Hopkins and Richardson, 88 slightly modified.

T. J. Holopainen / Vivarium 45 (2007) 1-29

These citations show that Anselm attributed to the single argument a very
central role in the argumentation of the Proslogion . In fact, he claims that
the Proslogion has been composed for the purpose of introducing the single
argument.

Ajiselm placed two requirements on the single argument that he was search¬
ing for—and we are justified in supposing that the argument he discovered in
his view fulfilled these requirements. First, the argument should be such that
it “would require no other [argument] than itself for proving itself”. The
significance of this requirement will be discussed in Part 4 below. The second
requirement concerns the issues that the single argument should be able to
prove. These issues can be conveniently divided into two items (as the transla¬
tors have done by adding the numbers (1) and (2) in the text). Item (1), prov¬
ing Gods existence, is rather clear. Anselm is obviously referring to his
argument for Gods existence in Proslogion 2 (or 2 and 3). Item (2) is not quite
as clear. What we are talking about are those things that are believed about the
Divine Substance or Essence. Because the notion that God is the Supreme
Good is part of that, item (2) can hence be summarized as follows: the single
argument should prove whatever is believed about the Divine Essence. But
what does this mean? What is one supposed to prove when one proves what¬
ever is believed about the Divine Essence? In the text of the Proslogion , there
are various kinds of statements related to the divine attributes and the other
properties of the Divine Essence, but can Ajiselm say this is all there is to be
believed about the Divine Essence? And how can Anselm say that a single
argument proves them, when he actually makes various kinds of argumenta¬
tive moves in the course of the Proslogion ?

To start from the last question, Anselm does not actually say that he will use
the single argument to prove whatever is believed about the Divine Essence.
He only promises to introduce an argument that can be used to do that. For
this reason, the focus of the analysis must be on the potential that some argu¬
ments have, and not so much on their actual use in the Proslogion.

The traditional theistic attributes are an important part of what is believed
about the Divine Essence. Anselm treats them in Monologion 15, where he
queries the predicates that substantially apply to the Supreme Being. He ends
the discussion by offering a systematization of the theistic attributes:

Just as it is blasphemous to suppose that the substance of the Supreme Nature is something
which in some respect it would be better not to be, so this substance must be whatever in
every respect it is better to be than not to be. For this substance alone is that than which
nothing at all is better; and it alone is better than all things which are not what it is....

T. J. Holopainen / Vivarium 45 (2007) 1-29

Therefore, necessarily, the Supreme Being is living, wise, powerful and all-powerful, true,
just, blessed, eternal, and whatever similarly is in every respect better than its negation.
Why, then, should I continue to ask what this Supreme Nature is, if whatever it is or is not
is evident? 9

In Anselm’s view, the Divine Essence is “whatever in every respect it is better
to be than not to be”, for example “living, wise, powerful and all-powerful,
true, just, blessed, eternal”. The same idea repeats itself in Proslogion 5, where
Anselm argues that God is “just, truthful, blessed, and whatever it is better to
be than not to be”. 10 If we name these kinds of attributes “great-making prop¬
erties”, we can say that in Anselms view the attributes of the Divine Essence
are great-making properties.

This systematization of the theistic attributes makes it possible to show that
that than which a greater cannot be thought must have any of the theistic
attributes. 11 For example, we can prove in the following way that that than
which a greater cannot be thought is wise:

If that than which a greater cannot be thought were not wise, it could be thought to be
wise, which is greater.

Therefore, if that than which a greater cannot be thought were not wise, then that than
which a greater cannot be thought would be that than which a greater can be thought.

But this is impossible.

Therefore, that than which a greater cannot be thought is wise.

This inference is modelled on the last part of Proslogion 2 and shares with it the
four features that were mentioned earlier. The inference is an indirect proof or
reductio ad absurdum. The contradiction which is derived is that that than

9) Anselm, Monologion 15, ed. Schmitt 1:29: “... sicut nefas est putare quod substantia supreme
naturae sit aliquid, quo melius sit aliquomodo non ipsum, sic necesse est ut sit quidquid omnino
melius est quam non ipsum. Ilia enim sola est qua penitus nihil est melius, et quae melior est
omnibus quae non sunt quod ipsa est.... Quare necesse est earn esse viventem, sapientem,
potentem et omnipotentem, veram, iustam, beatam, aeternam, et quidquid similiter absolute
melius est quam non ipsum. Quid ergo quaeratur amplius quid summa ilia sit natura, si manifes-
tum est quid omnium sit aut quid non sit?” Trans. Hopkins and Richardson, 26.

10) Anselm, Proslogion 5, ed. Schmitt 1:104: “Quod ergo bonum deest summo bono, per quod
est omne bonum? Tu es itaque iustus, verax, beatus, et quidquid melius est esse quam non esse.
Melius namque est esse iustum quam non iustum, beatum quam non beatum.”

10 See also Leftow, “Anselms Perfect-Being Theology”.

T. J. Holopainen / Vivarium 45 (2007) 1-29

which a greater cannot be thought is not what it is said to be. The contradic¬
tion is deduced by making a comparison between that than which a greater
cannot be thought as having a certain attribute and not having this attribute.
The comparison is made in terms of greatness. Because the traditional theistic
attributes are great-making properties, the kind of inference can successfully
be applied to any of them.

Anselm also applies the technique to some predicates that are not included
in the traditional list of divine attributes. An example of this is “not being only
in the understanding”, which is treated in Proslogion 2. This example also
shows that the predicates need not be linguistically simple but can be quite
complicated. But can we affirm that one can in this way prove “whatever is
believed of the Divine Essence”, as Anselm requires of his single argument?

In Anselms view we can. This is clear on the basis of an important passage
in the Responsio , viz. Responsio 10, which also otherwise confirms the view
about the functioning of the single argument that has been sketched. Anselms
main concern in the Responsio is to answer the objections presented by Gaunilo
and to point out some mistakes in Gaunilo s reading of his text. In the last
section, Responsio 10, Anselm, however, offers a more general comment on the
Proslogion. Given that Anselms outspoken objective with the publication of
the treatise was the introduction of the single argument, it is natural that he
comments on this argument in the concluding remark.

The first lines of Responsio 10 contain Anselms judgment on the validity of
his argument for Gods existence in Proslogion 2 (or Proslogion 2-3) and on the
force of the criticisms presented by Gaunilo.

I have now showed, I believe, that in the aforementioned treatise [viz., the Proslogion ] I
proved—not by inconclusive reasoning but by very compelling reasoning—that something
than which a greater cannot be thought exists in reality. And [I have showed] that this
[reasoning] was not weakened by any strong objection. 12

In the following lines, however, the perspective is widened. The remark which
follows is not only about a proof dealing with existence but also about proofs
dealing with whatever is believed about the Divine Essence.

12) Anselm, Responsio (= Quid ad haec respondeat editor ipsius libelli) 10, ed. Schmitt 1:138:
“Puto quia monstravi me non infirma sed satis necessaria argumentation probasse in praefato
libello re ipsa existere aliquid, quo maius cogitari non possit; nec earn alicuius obiectionis
infirmari firmitate.” Trans. Hopkins and Richardson, 130.

T. J. Holopainen / Vivarium 45 (2007) 1-29

For the signification of this utterance [viz., “something than which a greater cannot be
thought”] contains so much force that what is spoken of is, by the very fact that it is under¬
stood or thought, necessarily proved to exist in reality and to be whatever ought to be
believed about the Divine Substance. For we believe about the Divine Substance whatever
can in every respect be thought of as better [for something] to be than not to be. For
example, it is better to be eternal than not to be eternal, better to be good than not to be
good—or, rather, to be goodness itself than not to be goodness itself. But that than which
something greater cannot be thought cannot fail to be anything of this kind. Therefore, it
is necessarily the case that that than which a greater cannot be thought is whatever ought
to be believed about the Divine Being. 13

In this passage, Anselm offers a condensed description of how his single argu¬
ment works. The starting point of the argument is the utterance “that than
which a greater cannot be thought”. The conclusion is that that than which a
greater cannot be thought exists in reality and is whatever ought to be believed
about the Divine Substance. Importantly, we have here Anselms explicit state¬
ment that “we believe about the Divine Substance whatever can in every
respect be thought of as better [for something] to be than not to be”. Anselm
therefore believed that the reductio ad absurdum technique, explained above,
can be successfully applied exactly to those predicates that are believed to be
true of the Divine Essence.

Anselm does not make explicit reference to the reductio ad absurdum
technique in Responsio 10. Instead, he claims that the conclusions follow
from understanding the “signification of this utterance [‘that than which a
greater cannot be thought’]”. Here, we should notice two things. First, once
you have learned to master the reductio ad absurdum technique, it will become
redundant to you: you do not need to go through the reductio , it suffices that
you recognize a property as a great-making property. Second, it can be said
that the reductio ad absurdum is also based on the signification of the utterance
“that than which a greater cannot be thought”, for the contradiction that you
deduce is that that than which a greater cannot be thought is not what it is

,3) Ibid., 138-9: “Tantam enim vim huius prolationis in se continet significatio, ut hoc ipsum
quod dicitur, ex necessitate eo ipso quod intelligitur vel cogitatur, et revera probetur existere, et
id ipsum esse quidquid de divina substantia oportet credere. Credimus namque de divina sub¬
stantia quidquid absolute cogitari potest melius esse quam non esse. Verbi gratia: melius est esse
aeternum quam non aeternum, bonum quam non bonum, immo bonitatem ipsam quam non
ipsam bonitatem. Nihil autem huiusmodi non esse potest quo maius aliquid cogitari non potest.
Necesse igitur est quo maius cogitari non potest esse quidquid de divina essentia credi oportet.”
(The punctuation has been amended in the last two sentences.) Trans. Hopkins and Richardson,
130-1.

T. J. Holopainen / Vivarium 45 (2007) 1-29

said to be. Anselm does not explicate the reductio technique in Responsio 10,
because it should already be all too familiar to his readers.

The outlook on the single argument and its functioning that has been
sketched above is not rare in the mainstream of Anselm studies of recent
decades. In the Anglo-American literature at least, there is wide agreement
that the single argument is closely connected to Anselms attempt to
“establish’ the existence and nature of God by spelling out what is involved
in the single notion that than which a greater cannot be though f , 14 However,
the (relative) consensus concerns only the main outline. There has been
little explicit discussion about what the expression “single argument” exactly
refers to. 15 In what follows, I will argue for a specific identification of the single
argument: Anselms single argument is the notion or expression “that than
which a greater cannot be thought”. 16 Before this can be properly done, how¬
ever, the background in the early medieval theory of argument needs to be
discussed.

2. The Boethian Theory of Argument

Even though only one of Anselms complete treatises, viz. De grammatico , is
dedicated to a treatment of logical issues, it is clear that Anselm was a compe¬
tent logician fully conversant with the discussions in that thriving field of
study. 17 The main source for the logic or dialectic of Anselms time were the
commentaries and textbooks composed by Boethius in the early sixth century.

14) The formulation is that of Hopkins and Richardson in Anselm of Canterbury , Vol. 1: Monolo-
gion; Proslogion; Debate with Gaunilo; and a Meditation on Human Redemption , trans. Jasper
Hopkins and Herbert Richardson (Toronto, 1974), 153 n. 29.

15) For what The Cambridge Companion to Anselm has to say about the matter, see above, n. 2.

16) In the German discussions, the identification of the single argument as “that than which a
greater cannot be thought” has been quite common. See, e.g., Karl Barth, Fides quaerens intel-
lectum. Anselms Beweis der Existenz Gottes im Zusammenhang seines theologischen Programms
(Zollikon, 2 1958), 12; F. S. Schmitt in: Anselm von Canterbury ; Proslogion. Untersuchungen.
Lateinisch-deutsche Ausgabe (Stuttgart, 1962), 47-50; Klaus Kienzler, Glauben und Denken bei
Anselm von Canterbury (Freiburg, 1981), 220-3, 275-7. See also Holopainen, Dialectic and
Theology , 135.

17) For a basic survey of dialectic in the eleventh century, see Osmund Lewry, “Boethian Logic
in the Medieval West”, in Boethius: His Life , Thought and Influence, ed. Margaret Gibson(Oxford,
1981), 90-134, here esp. 94-108. For Anselms contribution as a logician, see Desmond Paul
Henry, The Logic of Saint Anselm (Oxford, 1967) and idem, Commentary on De grammatico*: The
Historical-Logical Dimensions of a Dialogue of St. Anselms (Dordrecht, 1974).

T. J. Holopainen / Vivarium 45 (2007) 1-29

Among Boethius’s works, two are dedicated to a branch of logic called “top¬
ics”. 18 When syllogistics focuses on the validity of the inference form, topics
focuses on discovering and confirming the premises in the inference. In its
Boethian form, the theory of topics assumes that there are self-evident univer¬
sal propositions, called “maximal propositions”, that can be used to prove
other propositions. The main part of topics consists of a discussion of various
maximal propositions, grouped under a number of headings called “diffe¬
rentiae”. Before presenting such a discussion, however, Boethius offers a gen¬
eral treatment of issues in the theory of argument. The following presentation
is based on Boethius’s treatment in Book One of In Ciceronis Topical

Topics was principally conceived as a heuristic technique for uncovering
arguments. In the text that Boethius was commenting on, Cicero compared
finding arguments to finding hidden things. You can find a hidden thing easily
if the place (locus in Latin, topos in Greek) where you should look for it is
indicated to you. In the same way, you can find arguments easily if you know
the “places” where it pays to look for them. 20 In topics, the places where argu¬
ments can be found are simply called “places”, loci\ the English technical term
is lopic .

The traditional Ciceronian definition of a “Topic” says that it is “the seat of
the argument” (sedes argumenti). According to Boethius’s account, the tradi¬
tion contains two competing ideas about what kinds of things Topics are: the
Aristotelian view says that maximal propositions are Topics, whereas the
Ciceronian view holds that differentiae of maximal propositions are Topics. 21

18) Boethius, In Ciceronis Topica , eds. J. C. Orelli and I. G. Baiter (Zurich, 1833), 270-388
(PL 64, 1039-1074); Boethius, De topicis differentiis , ed. Dimitrios Z. Nikitas (Athens, 1990);
PL 64, 1173-1218. For a general discussion of Boethius’s works on the topics, see Niels Jorgen
Green-Pedersen, The Tradition of the Topics in the Middle Ages: The Commentaries on Aristotle's and
Boethius 'Topics (Munich, 1984), 39-82.

,9) Cf. Holopainen, Dialectic and Theology , 135-9.

20) Cicero, Topica 2, 7, ed. G. Friedrich (Leipzig, 1912), 426: “Ut igitur earum rerum, quae
absconditae sunt, demonstrate et notato loco facilis inventio est, sic, cum pervestigare argumentum
aliquod volumus, locos nosse debemus; sic enim appellatae ab Aristotele sunt eas quasi sedes, e
quibus argumenta promuntur.” See also Green-Pedersen, Tradition of the Topics in the Middle
Ages, 44.

21) Boethius, In Ciceronis Topica I, [2, 7-2, 8], ed. Orelli and Baiter, 282-3: “Ex his etiam, quae
superius dicta sunt, quid distent Topica Ciceronis atque Aristotelis, apparuit. Aristoteles namque
de maximis propositionibus disserit: has enim locos argumentorum esse posuit, ut nos quoque
supra retulimus. Tullius vero locos non maximas propositiones, sed earum continentes differentias
vocat ac de his docere contendit.”

T. J. Holopainen / Vivarium 45 (2007) 1-29

As mentioned, maximal propositions are universal self-evident propositions.
Boethius explains the notion of maximal proposition ( maxima propositio ) as
follows:

We call highest and maximal propositions those propositions that are universal and known
and manifest to such an extent that they need no proof but rather themselves provide proof
for things that are in doubt, for those propositions that are undoubted are generally the
principles of demonstration for those propositions that are uncertain. Propositions of this
sort are ‘Every number is either even or odd’ and ‘If equals are subtracted from equals,
equals remain,’ and others whose truth is known and unquestioned. 22

The differentiae, for their part, are the headings under which the maximal
propositions are grouped. For example, there are some maximal propositions
that are related to definitions, and these are situated under the differentia
“from the definition” (a definitione) or “from the whole” {a toto). Correspond¬
ingly, those maximal propositions that concern the genus are situated under
the differentia “from the genus” ( a genere ), and so on. 23 From the practical
point of view, it makes little difference whether you use the term “Topic” to
refer to maximal propositions or to the differentiae of the maximal proposi¬
tions.

“Argument” (argumentum) was defined by Cicero as “a reason that produces
belief regarding a thing in doubt” (ratio rei dubiae faciens fidem). 2A As Boethius
construed this definition, the central expression in it is “thing in doubt” ( res
dubia)> which he understood as a technical term. Boethius starts from the
assumption that arguments are produced for the purpose of solving “ques¬
tions”. Also “question” ( quaestio ) is here a technical term, and it is defined as

22) Ibid., 280: “Supremas igitur ac maximas propositiones vocamus, quae et universales sunt et
ita notae atque manifestae, ut probatione non egeant, eaque potius, quae in dubitatione sunt,
probent. Nam quae indubitata sunt, ambiguorum demonstrationis solent esse principia, qualis
est: Omnem numerum vel parem esse vel imparem\ et: Aeqmlia relinqui, si aequalibus aequalia
detrahantur. ceteraeque, de quarum nota veritate non quaeritur.” Trans. Eleonore Stump in
Boethius’s In Ciceronis Topica (Ithaca, 1988), 33.

23) See ibid., 281.

24) Cicero, Topica 2, 8, ed. Friedrich, 426: “Itaque licet definire locum esse argumenti sedem,
argumentum autem rationem, quae rei dubiae faciat fidem.” Boethius, In Ciceronis Topica I,

[2,7-2, 8], ed. Orelli and Baiter, 276-7: “Argumentum ratio est, quae rei dubiae faciat fidem_

Addita igitur alia differentia, quae est rei dubiae , facta est integra definitio argumenti ex generee
duabus differentiis constans: genere quidem, ratione: una vero differentia, quod faciat fidem,
altera vero, quod rei dubiae, ut sit tota definitio, id esse argumentum, quod sit ratio rei dubiae
faciens fidem.”

T. J. Holopainen / Vivarium 45 (2007) 1-29

“a proposition in doubt” ( dubitabilispropositio). Not all interrogative sentences
are questions in this sense. A question asks whether or not something is the
case, for example “Is heaven spherical?”. Boethius claims that a question in a
way includes a contradiction: it contains an affirmation and a negation. For
example, the question “Is heaven spherical?” contains the affirmation “Heaven
is spherical” and the negation “Heaven is not spherical”. These two proposi¬
tions are called by Boethius the “parts” ( partes ) of a question. Now, the term
“thing in doubt” ( res dubia) is defined as “a part of a question” ( pars quaestionis ).
The expression “thing in doubt”, hence, can be used to refer to the affirmation
and negation that are contained in a question. 25 But when “a thing in doubt”
is understood in this way, the definition of “argument” will say that an argu¬
ment is a “reason” that produces belief regarding either the affirmation or the
negation that are contained in a question:

The whole purpose of an argument is directed toward a question, that is, toward a proposi¬
tion that is in doubt—not in order to prove the whole question but rather to corroborate
by reason a part of it, for one does not maintain a whole question but rather supports some
part of it by argumentation.... So since every question has two parts, an affirmation and a
negation, what one maintains is necessarily always based on one or the other part. One
person maintains the part that is an affirmation and another person the part that is a nega¬
tion, and each person seeks whatever arguments he can find, the first for support (ad astru -
endarri) of the affirmation and the second for its destruction (ad destruendam ). For it makes
no difference whether someone asserts an affirmation or destroys a negation, whether he
maintains a negation or opposes an affirmation. 26

What kind of entity should we assume an “argument” to be? The definition of
“argument” is not very helpful in this respect: all that it says is that an argu¬
ment is “a reason” (ratio). Boethiuss characterization of the term “argumenta¬
tion” (argumentatio) suggests one possible way of understanding what kind of
thing an argument is:

25) See Boethius, ibid., 277.

26) Ibid., 277-8: “Ad quaestionem igitur, id est, ad dubitabilem propositionem, omnis intentio

dirigitur argumenti, non vero ut totam comprobet quaestionem, sed ut partem eius ratione con¬
firmed neque enim tota quaestio defenditur, sed una eius quaelibet pars argumentatione
firmatur_Quum igitur omnis quaestio duas habeat partes, affirmationis unam, alteram nega¬

tion^, necesse est, ut sit semper ex alterutra parte defensio, ut unus quidem affirmationis partem,
negationis alter defendat, et hie quidem ad astruendam affirmationem, ille vero ad destruendam
quae potuerit argumenta perquirat. Nihil vero interest, utrum quis affirmationem ponat an
destruat negationem, aut negationem defendat an oppugnet affirmationem.” Trans. Stump, 30.

T. J. Holopainen / Vivarium 45 (2007) 1-29

But an argument will not be able to produce belief for something in doubt unless it is
expressed in speech and arranged with the interweaving of propositions. The expression and
arrangement of an argument by means of propositions is called an argumentation, and this
is said to be an enthymeme or a syllogism. 27

Since argumentation is the expressing of an argument, the argument can be
understood to be that which is expressed in the argumentation, or, as Boethius
formulates a little later, “the thought and meaning of the syllogism”. Accord¬
ing to this usage, then, a syllogism should not be called an argument. A syl¬
logism is an argumentation; the argument is that which the syllogism is used
to express. The matter is not this simple, however, because other passages in
In Ciceronis Topica suggest different ideas about what “arguments” are.

It was mentioned that Boethius describes it as Ciceros view that the
differentiae of maximal propositions are Topics. As part of his exposition
of the Ciceronian view, Boethius points out four possible ways of interpreting
the terms “argumentation” and “argument” and argues that in all four cases
the differentia of maximal propositions is “the seat of the argument”. 28 The
first three possible interpretations for the term “argument” are the following:

1. the argument is “the thought and meaning of the syllogism”

2. the argument is “the expression of the reasoning together with the maximal propositions
and the meaning of the syllogism”

3. the argument is “the maximal proposition”.

(In the fourth alternative, the term “argument” is interpreted in the same way
as in the first; the difference between these two alternatives is in the signification

27) Ibid., 278: “Argumentum vero, nisi sit oratione prolatum et propositionum contextione dis-
positum, fidem facere dubitationi non poterit. Ergo ilia per propositiones prolatio ac dispositio
argumenti argumentatio nuncupatur, quae dicitur enthymema vel syllogismus...” Trans. Stump,
31. See also Green-Pedersen, Tradition of the Topics in the Middle Ages, 44.

28) Ibid., 282: “Nam ex his quattuor significationibus appellationum duarum, argumentationis
scilicet atque argumenti, unam quamlibet esse necesse est. Aut enim elocutio et contextio ipsa
propositionum cum maximis propositionibus vel extra syllogismum positis vel in eodem inclusis
argumentatio vocatur; argumentum vero mens et sententia syllogismi. Aut elocutio ratiocinatio-
nis cum maximis propositionibus et sententia syllogismi argumentum esse dicetur, ut idem sit
argumentum, quod argumentatio. Aut argumentatio quidem vocabitur tota contextio syllogismi
cum sententia, sed argumentum maxima propositio. Aut integer ratiocinationis ordo praeter
maximas propositiones argumentatio dicetur, sententia vero argumentationis argumentum, reli-
qua vero maxima propositio locus. Sed quum haec ita sint... nihilo minus locos intelligimus
maximarum propositionum differentias....”

T. J. Holopainen / Vivarium 45 (2007) 1-29

of the term “argumentation”.) On the basis of the comments that Boethius
makes earlier, it appears that the first alternative is his preferred view, but he
fails to say it in this context.

Still, the signification of the term “argument” that is relevant for Anselm’s
argument is none of the three explicated by Boethius. Some passages in In
Ciceronis Topica suggest a fourth alternative: the term “argument” can be used
to refer to the middle term of the syllogism.

To appreciate the idea that a middle term can be called an argument, it is
important to recognize, first of all, the central role of terms in early medieval
dialectic (or in Aristotelian logic). The dialectical sentence analysis starts from
the analysis of a simple sentence into terms: a simple sentence consists of two
terms, the subject and the predicate. In categorical syllogistics, terms have a
constitutive role. A categorical syllogism consists of three simple sentences,
which contain three different terms (S, P and M), each of which appears in two
sentences: the predicate of the conclusion (P) appears also in the first premise;
the subject of the conclusion (S) appears also in the second premise; the third
term is the middle term (M) which appears in both of the premises. The
classification of valid syllogisms starts from a division of them into four figures,
on the basis of how the terms S, P and M are situated in the premises. 29

The idea that there is at least a very close connection between an “argu¬
ment” and the middle term of a syllogism comes up in a passage where
Boethius explains how a syllogism functions. The “question” that Boethius
uses as an example in this passage is “Whether a man is a substance or not”.
Boethius chooses to defend the affirmative part of the question, i.e., the part
“Man is a substance”. So this is the “thing in doubt” that he aims at confirming
with the aid of an argument, and at the same time the conclusion of the syl¬
logism that he will construct. The subject term of this question (and of the
“thing in doubt” or the “conclusion” of the syllogism to be constructed) is
“man”, and the predicate term is “substance”. From the syllogistic point of
view, the predicate term “substance” can also be called “the greater term”, and
the subject term “man” can be called “the lesser term”, and the both of them
can be called “the extreme terms” or “extremes” ( extremi ). To be able to con¬
struct a syllogism that proves his conclusion, Boethius needs to find a middle

29) The main sources for categorical syllogistics were Boethius’s textbooks Introductio ad syllogis-
mos categoricos (PL 64, 761-94) and De syllogismis categoricis (PL 64, 793-832). In the early
middle ages, the letters S, P and M were not yet used to refer to the three terms in the syllogism.
Instead, theexpressions “the lesser term” (= S), “the greater term” (= P) and “the middle term”
or “the common term” (= M) were used.

T. J. Holopainen / Vivarium 45 (2007) 1-29

term (medius terminus). The middle term that he will use is “animal”. He
explains the procedure as follows:

Since every syllogism consists of propositions, and propositions consist of terms, and the
terms differ from one another in that one is greater and the other is lesser, a conclusion
cannot arise from propositions unless the propositions, which proceed by means of terms,
conjoin the extreme terms by the intermediary of a third term. This is easily demonstrated
with an example. Suppose there is a question whether a man is a substance or not. I adopt
one part of the question in need of proof, namely, ‘Man is a substance.’ In this part there
are two terms, substance and man\ of these, substance is the greater and man is the lesser....
So in order for us to join man and substance, we must find a middle term that might unite
both terms. Let this be animal and let this be one premise: ‘Every man is an animal.’ In this
proposition animal is the predicate, and man is the subject. Then I add ‘But every animal is
a substance.’ In this proposition animal is now the subject and substance is the predicate.
And in this way I conclude, ‘Every man is a substance.’... Thus the extreme terms are
united by insertion of a middle term, and in this way the members of the question are
coupled with each other and the doubt is resolved by the proof employed. Hence an argu¬
ment is nothing other than the discovery of an intermediate, for an intermediate will be
able to conjoin the extremes, if an affirmation is being maintained, or to disjoin them, if a
negation is being asserted. 30

It is noteworthy that Boethius describes the syllogistic procedure also from the
point of view of terms, and from this point of view, of course, the middle term
will play the central role. In Boethius s description, the middle term is a term
that “unites both terms” (qui utrosque copulet terminos ), and when a syllogism
is formulated with the aid of it, “the extreme terms are united ( copulantur) by
insertion of a middle term”. In this way the middle term “will be able to con¬
join the extremes”, in the case where you argue for the affirmative part of the

30) Boethius, In Ciceronis Topica I, [2, 7-2, 8], ed. Orelli and Baiter, 279: “Quoniam igitur
syllogismus omnis propositionibus constat, propositiones vero terminis, terminique inter se
differunt eo, quod unus maior est, alter minor; fieri non potest, ut ex propositionibus conclusio
nascatur, nisi per terminos progressae propositiones extremos terminos alicuius tertii medietate
coniunxerint. Id facillimo demonstratur exemplo. Sit enim quaestio, utrum homo substantia sit,
an minime. Sumo mihi quaestionis partem alteram comprobandam, ea est, hominem esse substan-
tiam\ in hac igitur duo sunt termini, substantia atque homo, quorum maior substantia, homo vero

minor_Ut igitur substantiam atque hominem iungamus, necesse est medium terminum

reperiri, qui utrosque copulet terminos; hie sit animal, fiatque una propositio: Omnis homo ani¬
mal est. In hac igitur propositione animal praedicatur, homo subiicitur. Rursus adiungo: Omne
autem animal substantia est. In hac rursus animal supponitur, substantia praedicatur. Itaque con-
cludo: Omnis homo substantia est _Quoniam igitur extremi termini medii interpositione copu¬

lantur, eoque modo quaestionis inter se membra conveniunt, adhibitaque probatione solvitur
dubitatio; nihil est aliud argumentum quam medietatis inventio: haec enim vel coniungere, si
affirmatio defendatur, vel disiungere, si negatio vindicetur, poterit extremos.” Trans. Stump, 32.

T. J. Holopainen / Vivarium 45 (2007) 1-29

question, “or to disjoin then”, if you argue for the negative part. The power of
the syllogism is, hence, based on the power of its middle term. The centrality
of the middle term for the argument appears also in Boethius’s statement that
“an argument is nothing other than the discovery of an intermediate” (nihil est
aliud argumentum quam medietatis inventio).

In the above passage, Boethius does not exactly say that the middle term
could be called an argument. However, there is a series of other passages in
which Boethius suggests that some terms could function as arguments, and,
given the analysis of the functioning of a syllogism that Boethius presents, the
natural conclusion would be that he means the middle term of the syllogism.
The idea that some terms can function as arguments appears in a number of
passages in which Boethius denies that the subject or the predicate of a ques¬
tion could by themselves be used as arguments to solve the question:

By themselves these terms can neither be nor furnish arguments; for if they as they are
could be arguments or furnish the material for arguments, they would leave no doubt in the
question. 31

By denying that the terms of a question could, in themselves, be arguments,
Boethius raises the possibility that some other terms could serve as arguments.
And if we choose to call some term an argument, the most obvious choice will
be the middle term. As a matter of fact, at least in the case of a categorical syl¬
logism it is the only possible choice, for there are three terms in a categorical
syllogism, and the middle term is the only one that is not also a term of the
question.

To end the discussion of the Boethian theory of argument, let me empha¬
size that he does not just incidentally deny that the subject or predicate
term of a question could be an argument. It is a recurring theme in his discus¬
sion of Ciceros examples of arguments from different Topics in Book One
of In Ciceronis Topica? 2 To Boethius, it is a rule of dialectic that a term in a
question cannot by itself function as an argument to solve the question in
which it occurs.

30 Ibid., [2, 8], 283-4: “Ipsum itaque, de quo agitur, nihil est nisi uterlibet eorum terminus, qui
in quaestione proponitur, sive praedicatus, sive etiam subiectus. Qui quidem termini per se
argumenta esse non possunt, neque vero per se argumenta praestare. Si enim ipsi, ut sunt, argu-
menta esse possent vel argumentorum praestare materiam, nullam in quaestione linquererent
dubitationem.” Trans. Stump, 37.

32) Ibid., [2, 9], 288: “Est enim quaestio, an iuris civilis scientia sit utilis. Hie igitur ius civile
supponitur, utilis scientia praedicatur. Quaeritur ergo, an id, quod praedicatur, vere possit adhae-
rere subiecto. Ipsum igitur ius civile non potero ad argumentum vocare; de eo enim quaestio

T. J. Holopainen / Vivarium 45 (2007) 1-29

3. Some Contemporary Testimonies

Before turning to Anselm’s single argument, I would like to draw attention
to some pieces of evidence which confirm the potential relevance of the
above considerations for the interpretation of Anselms argument. It has
been observed that the Boethian sources do not contain any one idea of what
kind of entity you should take an argument to be, but they leave room for
discussion about the matter. Peter Abelard’s Super Topica glossae (from approx¬
imately 1115-20) confirms that there indeed was such discussion. Abelard
mentions three possible interpretations for the term “argument” that have
been advanced. The first view, which is also Abelard’s view, is that the term
“argument” refers to the propositions preceding the conclusion, i.e., the prem¬
ises in the argumentation. Some other people say that the argument is not
the premises but the meaning ( intellectus ) of these premises. 33 Thirdly, some
dialecticians

call neither the propositions nor their meaning “the argument”, but instead those things or
terms in the preceding propositions—as that in which the probative force resides—which
we call “Topics”. For example, when we say that Socrates is a man, wherefore he is an ani¬
mal, they call “man”—which is the Topic—“the argument”. 34

constituta est.” Ibid., [2, 10], 289-90: “Quaestio est igitur in proposito Ciceronis exemplo argu-
menti a partium enumeratione deducti: An is, quem servum fuisse constitit, liber sit. Is, quem ser-
vum fuisse constitit, subiectus est terminus, liber vero praedicatus. Neutrum igitur eorum
terminum ad argumentum ducere poterimus: de quibus enim dubitatur, ipsa fidem dubitationi
facere non possunt.” Ibid., [2, 10], 291: “Quaeritur, utrum, quum lex Aelia Sentia vindicem velit
esse assiduo assiduum, locupletem velit locupleti. Hie igitur subiectus quidem terminus est lex
Aelia Sentia vindicem volens assiduo assiduum, praedicatus vero locupletem locupleti. Ipsos igitur
terminos non potero ad fidem quaestionis adducere: de ipsis enim, de quibus ambigitur, nulla
effici fides potest.”

33) Peter Abelard, Super Topica glossae, in Scritti di logica, ed. Mario Dal Pra (Florence, 1969),
294: “Argumenti nomen tribus modis accipi potest. Modo autem argumentum nomen est prop-
ositionum quae praemittuntur ad probationem alterius propositionis, sive una praemittatur sicut
in enthymemate vel exemplo, sive plures sicut in sillogismo... Quidam vero non ipsas proposi¬
tions, sed earum intellectus argumentum vocant dicentes id scilicet recte argumentum dici in
quo vis est probationis, hoc est fidei constituendae, quod est intellectus....” For the doctrine of
the topics in the tenth to twelfth centuries, see Green-Pedersen, Tradition of the Topics in the
Middle Ages, 139-221, here esp. 171-2.

34) Peter Abelard, ibid., 294: “Alterum vero argumentum vocant proprie neque ipsas propositio¬
ns, neque earum intellectus, sed eas res vel eos terminos propositionum praecedentium, ut in
quibus vis est probandi, quod nos appellamus locos; veluti cum dicimus: Socrates est homo, quare
est animal, ipsum hominem qui locus est argumentum nominant...”

T. J. Holopainen / Vivarium 45 (2007) 1-29

We need not care here about the details of Abelards understanding of topics.
What concerns us is that Abelard says that some people call a thing or a term
“an argument”. In addition, it is remarkable that Abelard points out that these
people appeal to the authority of Boethius for their view:

Also Boethius himself, when he discusses the argument of syllogism in Book One of
his Commentary on Ciceros Topics , appears to call the middle term “the argument”. 35

After this, Abelard reproduces some passages from Book One of Boethius’s In
Ciceronis Topical Abelard’s remarks show that at least in the beginning of the
twelfth century it was a genuine option for dialecticians to interpret Boethius’s
In Ciceronis Topica as suggesting that the middle term is the argument.

Second, Anselm’s teacher, Lanfranc, makes routine use of topics in his com¬
mentaries to the Pauline Epistles, which shows that a familiarity with topics
was assumed in his school. 37 Also, in Lanfranc’s De corpore et sanguine Domini
(ca. 1063) there is a passage which makes use of a large number of terms and
principles that pertain to the theory of topics:

Hence above, where you [viz., Berengar] wanted to prove that the bread and the wine of the
altar do not undergo an essential change in the consecration, you adopted two [sentences]
as Topics for arguments, of which I proved with the aid of manifest reasons that one was
yours only, and the other was no-one’s. Here, you made a grave error. For what was yours
was the question. It is this that we are querying about: we endeavour to tear down and
smash it with all the weight and impact of arguments. Moreover, no question can be a Topic
for an argument. A Topic for an argument has to be either certain in itself or proved by
means of certain grounds. Therefore, what was only yours, should not at all have been
adopted to prove a thing in doubt. 38

35) Ibid., 295: “Ipse etiam Boetius in primo super Topica Ciceronis, cum de argumento sillogismi
loqueretur, ipsum medium terminum visus est argumentum vocare...”

36) Ibid., 295-6: “... Et rursus: ‘Quoniam igitur sillogismus omnis propositionibus constat,
propositiones terminis, termini inter se differunt, eo quod unus maior est, alter minor, fierinon
potest ut ex propositionibus conclusio nascatur, nisi per terminos progressae quaestionis extre-
mos terminos alicuius propositionis mediante termino coniunxerint.’ Et iterum: ‘Quoniam igi¬
tur, medii termini interpositione, extremi copulantur, eoque modo membra quaestionis inter se
conveniunt, adhibitaque probatione solvitur dubitatio, nihil est aliud argumentum quam medi-
etatis inventio;...’” Cf. Boethius, In Topica Ciceronis I, [2, 7-2, 8], ed. Orelli and Baiter, 279; see
above, n. 30.

37) See Margaret Gibson, “Lanfranc’s ‘Commentary on the Pauline Epistles’”, Journal of Theo¬
logical Studies , New Series, 22 (1978), 86-112, at 104-5.

38) Lanfranc of Canterbury, De corpore et sanguine Domini 7, PL 150, 417B-C: “Igitur superius
volens astruere panem vinumque altaris inter sacrandum essentialiter non mutari, duo quaedam

T. J. Holopainen / Vivarium 45 (2007) 1-29

Among the basic terms of topics that Lanfranc introduces into the text are the
following: Topic, argument, argumentation, question, and, a thing in doubt.
Lanfranc also claims that a Topic for an argument has to be either certain in
itself or proved by means of necessary grounds.

Thirdly, even though Anselm makes little use of terminology pertaining to
topics, also outside the Proslogion there is some evidence to show that he is
familiar with the kind of approach sketched above. In De grammatico 1, the
Student describes the question discussed in the work with terms reminiscent
of some passages in the sources of topics:

Furthermore, it is necessary that (an) expert-in-grammar be either a substance or a quality.
Thus, whichever one of these it is, it is not the other; and whichever one it is not, it has to
be the other. Accordingly, whatever suffices to prove (valet ad astruendam ) the one alterna¬
tive ( unam partem) disproves (destruit) the other; and whatever counts against the one
(debilitat) counts for the other ( roborat ). 39

In De grammatico 3, Anselm makes reference to the same distinction between
two kinds of acceptable premises that Lanfranc had referred to in De corpore et
sanguine Domini'.

Yet, I do not see that the foregoing premises are in any respect untenable. For the two
premises which have “man” as their subject-term are so self-evident (sic suntper se not*) that
it would be impudent to prove them; and the two premises which have “animal” as their

pro argumentorum locis assumpsisti, quorum unum tantummodo esse tuum, alterum nullius
hominum manifestis rationibus approbavi. In qua re magno vitio rem praedictam effecisti. Nam
quod tuum erat, quaestio erat. Ex eo quippe quaerimus id opprimere atque evertere omni argu¬
mentorum mole, atque impulsu satagimus. Porro nulla quasstio locus esse poterit argumenti.
Argumenti quippe locum necesse est, aut per se esse certum, aut certis rationibus approbatum.
Quod ergo tantummodo tuum erat, ad probandam rem dubiam assumi minime oportebat.”
Trans. Holopainen in Holopainen, Dialectic and Theology , 53-4. For Anselms relationship to
Lanfrancs De corpore et sanguine Domini , see Toivo J. Holopainen, “Logic and Theology in the
Eleventh Century: Anselm and Lanfranc s Heritage”, in Anselm and Abelard: Investigations and
Juxtapositions , eds. G. E. M. Gasper and H. Kohlenberger (Toronto, 2006), 1-16.

39) Anselm, De grammatico 1, ed. Schmitt 1:146: “Item quoniam necesse est ut grammaticus sit
aut substantia aut qualitas, ut quodlibet horum sit alterum non sit, et quodlibet non sit alterum
necesse sit esse: quidquid valet ad astruendam unam partem, destruit alteram, et quidquid unam
debilitat, alteram roborat.” Trans. Hopkins and Richardson, 132-3. The verbs astruere and
destruere are used as a pair in Boethius’s In Ciceronis Topica I, [2, 7-2, 8], ed. Orelli and Baiter,
278 (see above, n. 26). The verb astruere appears also in Anselm’s description of the single argu¬
ment: Proslogion, Prooemium, ed. Schmitt I, 93: “... unum argumentum, quod... solum ad
astruendum... sufficeret.” See above, n. 7.

T. J. Holopainen / Vivarium 45 (2007) 1-29

subject-term seem to be so well-established (sic videnturprobata ) that it would be impudent
to deny them. 40

In De grammatico 4, Anselm emphasizes the centrality of the middle term
(here called “a common term”) in an inference:

T. See, then, whether they [i.e., two sentences intended as premises] have a common term,
without which they do not entail any conclusion.

S. I see that they do not have a common term and, hence, that no conclusion follows from
them.

T. Does it seem to you, then, that in the case of these conjoined premises of yours no con¬
clusion at all can be inferred?

5. I certainly thought so. But your question makes me suspect that perhaps some logi¬
cal power lies hidden in them. Yet, without a common term how can they entail any
conclusion?

T. The common term of a syllogism must be common not so much in verbal form as in
meaning. For just as no conclusion follows if it is common in verbal form but not in mean¬
ing, so no harm is done if it is common in meaning but not in verbal form. Indeed, the
meaning—rather than the words—determines a syllogism. {Sententia quippe ligat syllogis-
mum, non verbal

This passage confirms that Anselm shared the view that what is central in
the argument is the middle term. There is also a methodological lesson to be
learned from the passage. While analysing the argumentative structure in the
Proslogion , there is no need to remain focused on the surface of the text, for
sententia ligat syllogismum , non verba.

40) Anselm, De grammatico 3, ed. Schmitt 1:147-8: “... licet prsecedentes propositiones in nullo
titubare videam. Duae namque quse subiectum terminum habent hominem, sic sunt per se notae,
ut imprudentia sit eas probare; duae vero quae subiciunt animal, sic videntur probatae, ut impu-
dentia sit eas negare.” Trans. Hopkins and Richardson, 135.

4,) Ibid., 4, 148-9: “M[agister]. Vide ergo utrum habeant communem terminum, sine quo nihil
efficiunt. D[iscipulus]. Video eas non habere communem terminum, et idcirco nihil ex eis conse-

qui_M. Itane tibi videtur his tuis conexionibus nihil concludi posse? D. Ita utique putabam,

sed haec tua interrogatio facit me suspectum, ne forte in illis aliqua lateat efficacia. Sed quomodo
efficiunt sine communi termino? M. Communis terminus syllogismi non tarn in prolatione
quam in sententia est habendus. Sicut enim nihil efficitur, si communis est in voce et non in
sensu: ita nihil obest, si est in intellectu et non in prolatione. Sententia quippe ligat syllogismum,
non verba.” Trans. Hopkins and Richardson, 136.

T. J. Holopainen / Vivarium 45 (2007) 1-29

4. An Argument Hiat Proves Itself

Anselms argumentation in the Proslogion is based on the power of the term
“that than which a greater cannot be thought”. In what follows, I will attempt
to show that Anselm looked at his argumentation from a Boethian standpoint
and that the term “that than which a greater cannot be thought” is his unum
argumentum . I will first suggest a way of analysing Anselms argumentation in
a “Boethian” manner, and then provide the textual evidence that supports this
kind of analysis. 42

One apparent difficulty in applying the Boethian approach to Anselms
argumentation is that Boethius treats the categorical syllogism as the paradigm
of argumentation, whereas Anselms argumentation makes use of a reductio ad
absurdum. This difficulty can be overcome, however. First, the most natural
reconstruction of Anselms argumentation includes categorical syllogisms in
which “that than which a greater cannot be thought” serves as a middle term.
Second, the Boethian approach can be applied also to the kind of reductio ad
absurdum that is involved here.

Anselm requires of his single argument that it ought to be able to prove a
number of sentences about God. If we look at these sentences from the point
of view of their terms, they all have the same subject term, viz. “God”, whereas
there are many different predicate terms, e.g. “wise”, “all-powerful”, “good”,
“goodness”, and so on. From the Boethian viewpoint, what Anselm needs in
order to be able to prove these sentences is a term that has the power of bring¬
ing together the extreme terms in each sentence. Further, it appears that
Anselm should be able to make use of the same middle term in all the cases in
order to present a single argument. As a matter of fact, several middle terms
can be suggested. The one that Anselm has in mind is, of course, “that than
which a greater cannot be thought”, which—according to Responsio 10—can
be proved both to exist and to be whatever the Divine Essence is believed to
be. Using this term as a middle term, it is possible to construct a categorical
syllogism for any of those sentences that the single argument should prove.
For example, we can prove that “God is wise” as follows:

God is that than which a greater cannot be thought
That than which a greater cannot be thought is wise
Therefore, God is wise.

42) Cf. Holopainen, Dialectic and Theology , 139-45.

T. J. Holopainen / Vivarium 45 (2007) 1-29

This kind of categorical syllogism is certainly not the most important thing
related to the single argument. As a matter of fact, Anselm does not spell out
any such syllogism in the Proslogion or in the Responsio. Nevertheless, the
idea is undeniably there. Anselm proves things of that than which a greater
cannot be thought in order that the conclusions can be applied to God {Pros¬
logion 2-3), and he proves things of God on the basis that He is that than
which a greater cannot be thought {Proslogion 5, 15, 18). A reconstruction of
such arguments with the aid of tools that were used in early medieval dialectic
will naturally (even though perhaps not inevitably) include a categorical
syllogism.

Using reductio ad absurdum inferences is a more important and more inter¬
esting part of Anselm’s argumentation. As shown in Part 1, Anselm can argue
for sentences like “That than which a greater cannot be thought is wise” by
presenting reductio ad absurdum inferences for their support. How should
such a reductio be analysed with the aid of early medieval tools? 43 If we apply
the approach that Boethius suggests in Book One of In Ciceronis Topica , we
should look for terms, i.e., expressions that function as subjects and predicates
in the sentences in the reductio. Because “that than which a greater cannot be
thought” and the predicate in question (say, “wise”) are the subject and predi¬
cate of the conclusion, they are to be treated as terms also in the preceding
sentences. If we leave them out, there is very little in the reductio that could
even potentially qualify as a subject or a predicate: there are expressions like
“can be thought” and “greater”, and none of these plays the central role that
the middle term is assumed to play. If we should name the most central term
in the reductio , there is only one reasonable alternative: the most central term
in the reductio is “that than which a greater cannot be thought”, on the
signification of which the reductio hinges. As Anselm says, the “signification of
this utterance” contains “so much force” that “what is spoken of is... necessar¬
ily proved to exist in reality and to be whatever ought to be believed about the
Divine Substance”. 44

A “Boethian” reconstruction of Anselm’s argument can, hence, be presented
as follows. You want to present a proof for a number of sentences about God.
You prove these sentences with the aid of categorical syllogisms in which “that

43) The Boethian dialectic does not contain explicit teaching on indirect proofs. In De syllogismis
categoricis the technique is used to prove some syllogistic moods with the aid of some other
moods, and to reduce some moods to some other moods, but the context is a special one. See
Boethius, De syllogismis categoricis II, PL 64, 818B-D and 820B-D.

44) See above, n. 13.

T. J. Holopainen / Vivarium 45 (2007) 1-29

than which a greater cannot be thought” is used as a middle term. To confirm
the major premises of these syllogisms, you again use the term “that than
which a greater cannot be thought” as a (quasi) middle term, but this time
not in a categorical syllogism but in an indirect proof. If we use the term
“argument” to refer to the middle term in the argumentation, we can say that
the same argument “that than which a greater cannot be thought” is used
twice in Anselms scheme of argumentation, first as a middle term and then
again as a (quasi) middle term to establish its own status as a middle term.

The preceding discussion shows that if we choose a particular way of inter¬
preting the early medieval theory of argument, we can construe Anselms argu¬
ment in the Proslogion in a certain way. But how can we know that Anselm
looked at his argument in this way? In the Proslogion proper, i.e., Ch. 1-26 of
the work, there appears to be no evidence either for or against this particular
interpretation. Nevertheless, there are two passages related to the Proslogion
which jointly confirm that the above analysis is at least close to Anselms own
way of looking at the matter. These are the passage in the Preface in which
Anselm introduces the single argument, and Section 5 of the Responsio.

In the Preface to the Proslogion , Anselm characterizes the single argument
by saying that it “would require no other [argument] than itself for proving
itself”. 43 Because Anselm is a very careful writer, we should start from the
assumption that every word counts in this characterization. Anselm does not
say that the single argument does not need any other argument for its support,
but that it needs “no other [argument] than itself for proving itself”. This is a
peculiar statement, which is possible to understand clearly only if we assume
that the word “argument” is used in some technical sense. As we have seen,
there is a perfectly good sense in which “that than which a greater cannot be
thought” can be characterized as an argument that can be used as an argument
to prove itself, for it can be used as a (quasi) middle term to establish its own
status as a middle term. Anselms remark that his argument “would require no
other [argument] than itself for proving itself”, hence, supports the view that
“that than which a greater cannot be thought” is Anselms argument, and it
also supports the “Boethian” analysis of the functioning of this argument.

Let us now turn to Responsio 5, which, combined with Anselms remark in
the Preface, contains clear evidence that the above reconstruction of the single
argument closely corresponds to Anselms own way of looking at the matter.

45) Anselm, Proslogion , Prooemium, ed. Schmitt 1:93: “... unum argumentum, quod nullo alio
ad se probandum quam se solo indigeret...” See above, n. 7.

T. J. Holopainen / Vivarium 45 (2007) 1-29

In Responsio 5 Anselm explicitly uses the term “argument” for notions like
“that than which a greater cannot be thought” or “the greater than all others”,
which shows that he was prepared to call terms “arguments”. 46 There is more,
however. As a matter of fact, the purpose of Responsio 5 is to point out the
difference between an argument that requires some other argument than itself
for proving itself and an argument that requires no other argument than itself
for proving itself. The argument “the greater than all others” is an example of
the former type, whereas the argument “that than which a greater cannot
be thought” is an example 1 —in Anselm’s view, the unique example—of the
latter type.

In Responsio 5, Anselm rebukes Gaunilo for mispresenting his argument. In
his paraphrases of Anselm’s argumentation in Proslogion 2, Gaunilo had repeat¬
edly used the notion “the greater than all others” (mains omnibus) instead of
the notion “that than which a greater cannot be thought”. In Anselm’s view,
this substitution makes the argumentation entirely different:

(...) you say repeatedly that I argue as follows: “That which is greater than all [others] is in
the understanding. And if it is in the understanding, it exists also in reality; for otherwise
[i.e., if it did not exist in reality, that which is] greater than all [others] would not be greater
than all [others].” But nowhere in all of my statements is there found such a line of reason¬
ing. For the expression “[that which is] greater than all [others]” and the expression “that
than which a greater cannot be thought” are not equally effective in proving that what is
spoken of exists in reality. 47

Here Anselm claims that there is an important difference between the notions
“the greater than all others” and “that than which a greater cannot be thought”,

46) Anselm, Responsio 5, ed. Schmitt 1:135: “An hie sic aperte inferri potest: non est ergo maius
omnibus quae sunt, sicut ibi apertissime diceretur: ergo non est quo maius cogitarine-quit? Illud
namque alio indiget argumento quam hoc quod dicitur ‘omnibus maius’; in isto veronon est
opus alio quam hoc ipso quod sonat ‘quo maius cogitari non possit’. Ergo si non similiter potest
probari de eo quod ‘maius omnibus’ dicitur, quod de se per seipsum probat ‘quo maius nequit
cogitari’: iniuste me reprehendisti dixisse quod non dixi, cum tantum differat ab eo quod dixi.
Si vero vel post aliud argumentum potest, nec sic me debuisti reprehendere dixisse quod probari
potest. Utrum autem possit, facile perpendit, qui hoc posse ‘quo maius cogitari nequit’ cognos-
cit.” Cf. below.

47) Ibid., 134: “Primum, quod saepe repetis me dicere, quia quod est maius omnibus est in intel-
lectu, si est in intellectu est et in re—aliter enim omnibus maius non esset omnibus maius—:
nusquam in omnibus dictis meis invenitur tabs probatio. Non enim idem valet quod dicitur
‘maius omnibus’ et ‘quo maius cogitari nequit’, ad probandum quia est in re quod dicitur.” Trans.
Hopkins and Richardson, 125.

T. J. Holopainen / Vivarium 45 (2007) 1-29

when it comes to proving the existence of the thing. The difference, we will
see, is that the one needs another argument for its support, whereas the other
does not.

Anselm begins by considering, once again, the notion “that than which a
greater cannot be thought”. If someone should claim that that than which a
greater cannot be thought does not exist, or that it is able not to exist, or that
it can be thought not to exist, this person can be easily refuted. Anselm thinks
that the three claims are connected to each other in such a way that it is
enough to prove the last point. This can be done by using the reductio ad
absurdum often cited above: if that than which a greater cannot be thought
can be thought not to be, it is not what it is said to be (which is, of course,
impossible). 48 But, Anselm claims, it seems that you cannot use such a simple
proof in the case of the greater than all others. If someone says that the greater
than all others can be thought not to exist, you are not entitled straightaway
to infer that it is not the greater than all others. Anselm argues that “the greater
than all others” cannot be used as an argument to prove itself but needs another
argument for its support:

However it seems that it is not as easy to prove this in respect of what is said to be greater
than all [others].... For what if someone should say that something that is greater than all
[others] actually exists, and yet that this same being can be thought of as not existing, and
that something greater than it can be thought, even if this does not exist? In this case can it
be inferred as evidently that [this being] is therefore not that which is greater than all [oth¬
ers], as it would quite evidently be said in the other case that it is therefore not that than
which a greater cannot be thought? In the former case another argument is needed {alio
indiget argumento) than the one that is uttered, “the greater than all [others]”; in the latter
case no other [argument] is needed {non est opus alio ) than the one that resounds, “that than
which a greater cannot be thought”. 49

48) Ibid., 134-5: “Si quis enim dicat quo maius cogitari non possit’ non esse aliquid in re
aut posse non esse aut vel non esse posse cogitari, facile refelli potest. Nam quod non est, potest
non esse; et quod non esse potest, cogitari potest non esse. Quidquid autem cogitari potest non
esse: si est, non est quo maius cogitari non possit. Quod si non est: utique si esset, non esset quo
maius non possit cogitari. Sed dici non potest, quia ‘quo maius non possit cogitari’ si est, non est
quo maius cogitari non possit; aut si esset, non esset quo non possit cogitari maius. Patet ergo
quia nec non est nec potest non esse aut cogitari non esse. Aliter enim si est, non est quod dicitur;
et si esset, non esset.”

49) Ibid., 135: “Hoc autem non tarn facile probari posse videtur de eo quod maius dicitur omni¬
bus .... Quid enim si quis dicat esse aliquid maius omnibus quae sunt, et idipsum tamen posse
cogitari non esse, et aliquid maius eo etiam si non sit, posse tamen cogitari? An hie sic aperte
inferri potest: non est ergo maius omnibus quae sunt, sicut ibi apertissime diceretur: ergo non est

T. J. Holopainen / Vivarium 45 (2007) 1-29

Consequently, Anselm claims that Gaunilos critique is misplaced. If the cri¬
tique is at all valid, it applies to Gaunilos version but does not apply to
Anselms:

Therefore, if what “that than which a greater cannot be thought” of itself proves concerning
itself (< de se per seipsum probat) cannot be proved in the same way in respect of what is said
to be “greater than all [others]”, you criticize me unjustly for having said what I did not say,
since it differs so much from what I did say. 50

However, Anselm also thinks that Gaunilos version can be saved, and this can
be done by using the notion “that than which a greater cannot be thought” as
an argument:

If, however, it can [be proved] by means of another argument ( post aliud argumentum
potest ), you should not have criticized me for having asserted what can be proved. Whether
it can [be proved], however, is easily appreciated by one who recognizes that “that than
which a greater cannot be thought” can do it {qui hoc posse quo maius cogitari nequit cognos-
cit). For one cannot in any way understand “that than which a greater cannot be thought”
without [understanding that it is] that which alone is greater than all [others]. Therefore,
just as that than which a greater cannot be thought is understood and is in the understand¬
ing and hence is affirmed to exist in reality, so what is said to be greater than all [others] is
inferred to be understood and to be in the understanding and, hence, necessarily, to exist
in reality. 51

In this passage, Anselm suggests that “that than which a greater cannot be
thought” can be used as an argument to prove that the greater than all others
exists in reality. Using the reductio ad absurdum , it is easy to prove that that

quo maius cogitari nequit? Illud namque alio indiget argumento quam hoc quod dicitur ‘omni¬
bus maius’; in isto vero non est opus alio quam hoc ipso quod sonat ‘quo maius
cogitari non possit’.” My translation on the basis of M.J. Charlesworth’stranslation in St. Anselm’s
Proslogion (Oxford, 1965), 181.

50) Ibid.: “Ergo si non similiter potest probari de eo quod ‘maius omnibus’ dicitur, quod de se
per seipsum probat ‘quo maius nequit cogitari’: iniuste me reprehendisti dixisse quod non dixi,
cumtantum differat ab eo quod dixi.” Trans. Charlesworth, 181 (slightly modified).

51) Ibid.: “Si vero vel post aliud argumentum potest, nec sic me debuisti reprehendere dixisse
quod probari potest. Utrum autem possit, facile perpendit, qui hoc posse ‘quo maius cogitari
nequit’ cognoscit. Nullatenus enim potest intelligi ‘quo maius cogitari non possit’ nisi id
quod solum omnibus est maius. Sicut ergo ‘quo maius cogitari nequit’ intelligitur et est in intel-
lectu, et ideo esse in rei veritate asseritur: sic quod maius dicitur omnibus, intelligi et esse in
intellectu, et idcirco re ipsa esse ex necessitate concluditur.” My translation (cf. Charlesworth’s
translation, 182-3).

T. J. Holopainen / Vivarium 45 (2007) 1-29

than which a greater cannot be thought is greater than all others, and therefore
the conclusions which apply to the former apply also to the latter. In this
way, “that than which a greater cannot be thought” can function as a middle
term that brings together the terms “the greater than all others” and “exists in
reality”.

Anselm ends Responsio 5 as follows:

You see, then, how right you were to compare me with that stupid person who wished
to maintain that the Lost Island existed from the sole fact that being described it was under¬
stood. 52

The significance of this ironical remark can be explained as follows. As part of
his criticism of Anselms argument for God’s existence, Gaunilo had suggested
the counterexample of a Lost Island. Anselm rejects this counterexample in
Responsio 3, but he is not too explicit about his grounds for doing so. What
Anselm states quite clearly is that he considers his argument to be unique:

With confidence I reply: if besides that than which a greater cannot be thought anyone
finds for me [anything else] (whether existing in reality or only in thought) to which he can
apply the logic of my argument, then I will find and will make him a present of that Lost
Island—no longer to be lost. 53

In this passage, Anselm claims that the technique of argumentation used by
him cannot be used in connection with any other notion than “that than
which a greater cannot be thought”. The remark at the end of Responsio 5
implies that Anselm connects the uniqueness of “that than which a greater
cannot be thought” to the feature that it requires no other argument than itself
for proving itself. Correspondingly, the “Lost Island” suffers from the same
deficiency as “the greater than all others”: if someone denies its existence, you
cannot use a reductio ad absurdum based on the notion “Lost Island” (or the
notion “the most excellent island”) to prove your case. (Anselm does not
explain why.)

52) Ibid., 135-6: “Vides ergo, quam recte me comparasti stulto illi, qui hoc solo quod descripta
intelligeretur, perditam insulam esse vellet asserere?” Trans. Charlesworth, 183.

53) Responsio 3, ed. Schmitt 1:133: “Fidens loquor, quia si quis invenerit mihi aut re ipsa aut sola
cogitatione existens praeter ‘quo maius cogitari non possit’, cui aptare valeat conexionem huius
meae argumentations: inveniam et dabo illi perditam insulam amplius non perdendam.” Trans.
Hopkins and Richardson, 123 (slightly modified).

T. J. Holopainen / Vivarium 45 (2007) 1-29

Anselms conviction that “that than which a greater cannot be thought” is
unique as an argument, in the sense that it can be used as an argument to
prove itself, appears to be related to some of the Boethian ideas discussed
above. Boethius presents it as a rule of dialectic that a term in a question can¬
not by itself function as an argument to solve the question in which it occurs. 54
Anselms technique appears to counter this rule, for he uses “that than which
a greater cannot be thought” as an argument to prove sentences in which “that
than which a greater cannot be thought” appears as the subject-term. Anselm
assumed that his argument would be the sole exception to the rule, and there¬
fore he had every reason to believe that his argument is unique. Hence, he
could confidently promise the Lost Island to the one who finds anything else
than “that than which a greater cannot be thought” to which the logic of his
argument would apply.

To recapitulate, Responsio 5 shows that “that than which a greater cannot be
thought” is for Anselm an “argument” (argumenturri ), and it is for him an
argument that “requires no other argument than itself for proving itself”. In
addition, Responsio 3 and Responsio 5 show that Anselm believed it to be the
only argument that requires no other argument than itself for proving itself. It
follows, therefore, that “that than which a greater cannot be thought” is
Anselms single argument. Responsio 5 also makes it obvious that the presented
“Boethian” reconstruction of the single argument at least approximately
corresponds to Anselms own outlook on his argument.

54) See above, n. 31 and 32.

BRILL

VIVA

RIUM

Vivarium 45 (2007) 30-68 www.brill.nl/viv

Albert the Great and the Revival of Aristode’s
Zoological Research Program

Michael W.Tkacz

Gonzaga University .; Spokane, Washington, USA

Abstract

Although Aristotle’s zoological works were known in antiquity and during the early
medieval period, the scientific research program discussed and exemplified therein
disappeared after Theophrastus. After some fifteen hundred years, it reappears in the
work of Albert the Great who extensively explains Aristotle’s conception of a scientific
research program and extends Aristotle’s zoological researches. Evidence of Albert’s
Aristotelian commentaries shows that he clearly understood animals to represent a self-
contained subject-genus, that the study of this subject-genus constitutes theoretical
knowledge in an Aristotelian sense, that natural finality and suppositional necessity
provide principles of zoological science, and that research into animals must be con¬
ducted according to a two-staged methodology of division and demonstration.

Keywords

Albertus Magnus, Aristotelian zoology, division, demonstration

I. Introduction

The year 1249 marked a turning point in the intellectual career of Albert the
Great, for this was the year in which he finally acceded to the pleas of his
Dominican confreres to compose a work explaining the natural science of
Aristotle. 1 This work, his paraphrastic commentary on the Physics , was to be
the first part of one of the major literary productions of the Middle Ages, a

0 Albert makes his intention clear toward the beginning of his commentary: “Intentio nostra in
scientia naturali est satisfacere pro nostra possibilitate fratribus ordinis nostri, nos rogantibus ex
pluribus iam praecedentibus annis, ut talem librum de physicis eis componeremus, in quo et
scientiam naturalem perfectam haberent et ex quo libros Aristotelis competenter intelligere pos-
sent.” Physica I, tr. 1, c. 1 in Alberti Magni Opera Omnia (Munster im Westfi, 1951-) [editio

DOI: 10.1163/156853407X195105

M. W. Tkacz / Vivarium 45 (2007) 30-68

production which would establish Albert as an auctoritas on equal footing
with Avicenna, Averroes, and Aristotle himself. 2 The significance of this event
is indicated by Albert’s own stated intention of making the new learning of
Aristotle intelligible to the Latins. 3 In fact, Alberts plan was far more ambi¬
tious than his Dominican brethrern could have imagined. He intended not
merely to produce an elementary guide to the Physics , nor just a systematic
commentary of the whole Aristotelian corpus. 4 Rather, he set out to reestab¬
lish the natural sciences by reviving the long dormant research programs of
Aristotle himself. 5

This intention is especially evident in Albert’s treatment of Aristotle’s
zoological works as well as his own biological studies. 6 Indeed, Albert is the
first scholar since Theophrastus to show any interest in Aristotelian zoology as
a research discipline. James G. Lennox, drawing attention to the disappear¬
ance of the particular sort of biological investigation that engaged the early

Coloniensis], 4/1:1.9-14. References to Alberts works are made according to the method
described in the American Catholic Philosophical Quarterly 70 (1996), 6-9. On the date of Alberts
Physica , see James A. Weisheipl, O.P. “Albert the Great and Medieval Culture,” The Thomist 44
(1980), 494; see also “Alberts Works on Natural Sciences ( libri naturales) in Probable Chrono¬
logical Order,” in Albertus Magnus and the Sciences: Commemorative Essays 1980 , ed. James A.
Weisheipl, O.P. (Toronto, 1980), 565 [1]. Cf. Paul Hossfeld, editor of the Physica for the editio
Coloniensis, who dates the work between 1248 and 1257 (ed. Colon. 4/1: prolegomena, v).

2) Roger Bacon testifies to this when he complains “nam sicut Aristoteles [et] Avicenna et Aver¬
roes allegantur in scholis, sic et ipse; et adhuc [Albertus] vivit et habuit in vita sua auctoritatem,
quod nunquam homo habuit in doctrina.” Opus Tertium, ed. J. S. Brewer (London, 1859), 30.

3) “... nostra intentio est omnes dictas partes [physicam et metaphysicam et mathematicam]
facere Latinis intelligibiles.” Physica I, tr. 1, c. 1 (ed. Colon. 4/1:1.48-49).

4) The manuscript evidence reveals the systematic and comprehensive character of Albert s com¬
mentaries. The autograph of Albert’s Aristotelian commentaries in the Osterreichische National-
bibliothek at Vienna (Cod. misc. Latin, 273) shows that the commentaries were written
consecutively in the order of the Aristotle’s books as then arranged. Folio 72 V contains the last
five lines of the Physica and continues straightaway on the same folio with the opening of the De
caelo. This later work ends and is immediately followed with the De natura locorum on folio l42 r .
See, Weisheipl, “The Life and Works of St. Albert the Great,” in Albertus Magnus and the Sciences ,
30-31. For a complete list of Albert’s Aristotelian commentaries and studies on the manuscript
tradition see Charles H. Lohr, “Medieval Latin Aristotle Commentaries, Authors A-F,” Traditio
23 (1967), 338-45.

5) Albert himself testifies to his intention to complete the Aristotelian corpus where necessary:
“Et addemus etiam alicubi partes librorum imperfectas et alicubi libros intermissos vel omissos,
quos vel Aristoteles non fecit vel forte si fecit, ad nos non pervenerunt.” Physica I, tr. 1, c. 1 (ed.
Colon. 4/1:1.38-41).

6) Albert commented twice on Aristotle’s zoological works. The first is a series of Quaestiones
super De animalibus disputed at the Dominican studium at Cologne in 1258 and preserved in
the reportatio of Conrad of Austria. The critical edition of this text produced by Ephrem Filthaut

M. W. Tkacz / Vivarium 45 (2007) 30-68

Peripatetics, argues that a mystery attends the origin of biological science. 7
The zoological treatises of Aristotle as well as the botanical studies of his stu¬
dent Theophrastus had set out a systematic program of theoretical research.
These works were known throughout late antiquity and, in the Greek and
Arabic east, during the early medieval period. Yet, as Lennox suggests, no
naturalist until Albert pursued or developed this research program.

While Lennox has established that Aristotelian zoology as a research pro¬
gram was not pursued in late antiquity, the evidence supporting his suggestion
regarding Albert has yet to be fully set out and analyzed. 8 Before it can be
established that Aristotelian zoology as a research program reappeared in the
work of Albert, the program must first be clearly defined and, second, it must
then be shown that Albert understood this research program in the same way
as the early Peripatetics. The first of these tasks has already been accomplished
by Lennox who articulates the Aristotelian notion of research program in the
form of four minimal requirements. 9 The first requirement is that such a pro¬
gram must identify a relatively self-contained domain as a single subject of
investigation. Second, a research program must be designed to obtain theo¬
retical knowledge of its subject domain. Third, fundamental principles for the
research program must be explicitly identified and articulated. Finally, a set of
domain-specific concepts and methods must be defined in such a way that the
research program is able to answer the questions that arise regarding the sub¬
ject domain.

The second task is accomplished by the present essay, which shows that
Aristotles conception of zoological research reappears, after a long dormancy,
in the work of Albert. For Albert’s careful and detailed commentaries on Aris¬
totle’s zoological treatises, as well as his general treatment of natural philoso-

is found in ed. Colon. 12: 77-309. The second is a paraphrastic commentary published together
with Albert’s original zoological researches in his massive De animalibus libri XXVI , edited from
the Cologne autograph by Hermann Stadler in Beitrdge zur Geschichte cLer Philosophic des Mit -
telalters (Munster, 1916 and 1920) [editio Stadler], Bd. 15-16. This latter text has been recently
translated as Albertus Magnus on Animals: A Medieval Summa Zoologica , tr. Kenneth F. Kitchell,
Jr. and Irven M. Resnick (Baltimore, 1999).

7) “The Disappearance of Aristotle’s Biology: A Hellenistic Mystery,” Apeiron 27 (1994), 7-24;
reprinted in Lennox’s Aristotle’s Philosophy of Biology: Studies in the Origins ofLife Science (Cam¬
bridge, 2001), 110-125. Page references are to this reprint edition.

8) In an “unsatisfactory postscript” to his study, Lennox does not provide evidence, citing only
my doctoral dissertation, The Use of the Aristotelian Methodology of Division and Demonstration in
the De animalibus of Albert the Great (The Catholic University of America, 1993), which treats
only some of the requirements of an Aristotelian research program. See Lennox, 123-24.

9) Lennox, 110-14.

M. W. Tkacz / Vivarium 45 (2007) 30-68

phy, articulate an Aristotelian research program precisely according to the four
criteria discussed by Lennox.

II. The Subject of Zoological Research

Toward the beginning of his commentary on Aristotle’s Physics , Albert notes
that every science has a generic subject matter. The task of the researcher work¬
ing in a science is to prove the properties of the subject matter and to investi¬
gate its specific parts. 10 Intellectual reflection on sense experience reveals the
subject for natural science, namely, changeable body (corpus mobile). Such
reflection identifies this subject matter by an analysis or “sorting out” process
that aims at a general definition, and the most general defining feature of sen¬
sible objects is their changeability. 11 Yet the same reflective analysis also reveals
that changeable body exists in many ways. A general science of nature, then,
can be divided into specific sub-sciences on the basis of subject matter. 12 In
particular, as changeable body can be either inanimate or animate, there is a
specific natural science of each. 13 Animate changeable body, however, can be
further divided into vegetative and sensitive. It is the sensitive animal that is
the subject of zoological research. Zoology, then, is a distinct science with a
distinct subject matter and, at the same time, it is a part of a general science of
nature because its subject is a species of changeable body. 14 Albert adds the
observation that this division of natural science is reflected in the relationship
of Aristotle’s various libri naturales . 15

10) “Omnis enim scientia est alicuius generis subiecti, de quo probantur passiones et cuius con-
siderat partes et differentias.” Physica I, tr. 1, c. 3 (ed. Colon. 4/1:3.23-23).
n) “Hoc autem in omni scientia naturali absque dubio est corpus mobile, prout motui subicitur.
Voco autem corpus subiectum physicae in communi et non hoc corpus vel illud. In communi
autem accipio non simpliciter, sed quod motui subicitur.” Physica I, tr. 1, c. 3 (ed. Colon.
4/1:3.25-30).

12) “In particularibus enim rebus omnis scientia particularis habet differentias notas, per quas
separatur ab alia particulari scientia.” De animalibus 72, tr. 1, c. 1 (ed. Stadler 15:762.36-38).

13) Physica I, tr. 1, c. 4 (ed. Colon. 4/1:6-8, esp. 7:8-64); De anima I, tr. 1, c. 1 (ed. Colon.
7/1:1-3); De animalibus 72, tr. 1, c. 1 (ed. Stadler 15:762.39-763.5).

14) Albert makes this point directly in discussing the subject of Aristotle’s zoology: “... sed ista
scientia [de animalibus] est pars scientiae naturalis; ergo subiectum huius erit pars subiecti totius;
sed subiectum totius est corpus mobile; ergo subiectum huius est corpus mobile specificatum.”
Quaestiones de animalibus I, q. 1 (ed. Colon. 12:77.34-38); see also De anima I, tr. 1, c. 1 (ed.
Colon. 7/1:1-2).

,5) Albert discusses this in detail in Physica I, tr. 1, c. 4 (ed. Colon. 4/1:6-8). In this text, after
indicating the books which deal with the various powers of the animating soul, he adds (ed.
Colon. 4/1:7.59-64):

M. W. Tkacz / Vivarium 45 (2007) 30-68

Another way of identifying the subject of a science is through a functional
analysis of what is given in sense perception. Living beings are distinguished
from non-living beings by analyzing the way in which their parts are formed
and exist for the sake of realizing certain capacities. 16 Albert identifies the most
basic parts of a living organism as those directed at nutrition, for the capacity
for nutrition is co-extensive with the life of the organism and that upon which
other universal and basic life functions, such as growth and reproduction,
depend. 17 It is clear from sense perception that not everything can be nour¬
ished, for nourishment requires a body with certain parts and these parts must
compose a living (ensouled) organism. Retention of life and growth are the
most obvious consequences of nutrition and these, Albert makes clear, are
properly said only of living organisms that possess a certain power he calls
“augmentative force” (vis augmentativa) , 18 Living organisms, then, comprise a
distinct type of subject defined by their composition of parts existing for the
sake of life functions.

Animals are a sub-type of living organism distinguished from plants by
their possession of the distinctive capacities of perception and, at least in most
cases, locomotion. At the same time, animals share with all living organisms
the basic life-functions of nutrition, growth, and reproduction. 19 Albert points
out that, of all the special disciplines of natural science, the study of animals
is last because

“Quibus habitis sufficit addere scientiam de corpore animato vegetabili et sensibili, cuius differ¬
entiae quoad vegetabilia traduntur in libris De vegetabilibus, et quoad differentiae animalium
traditur scientia sufficiens in libris De animalibus. Et ille liber est finis scientiae naturalis.”

16) “Et eadem de causa dicimus, quod creatio corporis animati est propter animam et opera
ipsius, et membra corporis creata fuerunt organica propter virtutes animae et operationes ipsius,
et natura et figura cuiuslibet membrorum est conveniens operationi ad quam est praeparata.” De
animalibus XI, tr. 2, c. 3 (ed. Stadler 13:793.10-14).

17) Deanima II, tr. 2, c. 2 (ed. Colon. 7/1:84v. 15-21); De generatione et corruptionel, tr. 3, c. 15
(ed. Colon. 5/2:150); De nutrimento et nutribili tr. 1, c. 5 and tr. 2, c. 1 in Bead Alberti Magni
Opera Omnia , ed. Auguste Borgnet (Paris: Vives, 1890-99) [ed. Borgnet], 9:333b-334a and
336a-b. On Alberts treatment of growth and nutrition see Joan Cadden, “Albertus Magnus’
Universal Physiology: the Example of Nutrition,” in Albertus Magnus and the Sciences: Com¬
memorative Essays 1980, ed. James A. Weisheipl (Toronto, 1980), 321-39.

,8) “Proprie quidem augetur, quod est animatum habens virtutem animae, quae dicitur vis aug¬
mentativa ...” De generatione et corruptionel, tr. 3. c. 11 (ed. Colon. 5/2:149.10-12).

19) Physica I, tr. 1, c. 4 (ed. Colon. 4/1:6-8, esp. 7.8-64); Quaestiones de animalibus XII, q. 5
(ed. Colon. 12:227-28); De animalibus I, tr. 1, c. 1 (ed. Stadler 15:1.1-12) and XI, tr. 1, c. 1
(ed. Stadler 15:762.39-763.5).

M. W Tkacz / Vivarium 45 (2007) 30-68

an animal is that which is more composite among natural things in the body and in the
soul. In the body it has the elements and their commixture, coagulation, complexion,
composition from heterogeneous materials, and communication with the soul of evident
life, which is the perfect soul. Other natural things either do not possess this composition
or do not possess it to this degree. Simple things do not possess it and minerals and plants
do not possess it to the degree that animals are composed. In the souls of animals there
is sensation and operation of evident life, which contain in themselves vegetation and the
operation of non-evident life, as the tetragon contains in itself the triangle. On account
of both of these, therefore, the science of animals ought to be last in the natural sciences
(in rebusphisicis). 20

Zoology has a subject-domain distinct from that of the other natural sciences
as well as from biology in general. Animals clearly manifest the basic life func¬
tions that plants do and nonetheless possess distinguishing capacities not pos¬
sessed by plants. Moreover, animals share both the common principles of life
as well as their distinctive animal functions despite their diversity of species.
The study of animal life, then, represents a unified field of investigation on the
basis of a functionally delineated subject of research. 21

Albert realized that identification of a subject-domain cannot simply be a
matter of delineating a genus of sensible objects to be studied. In his com¬
mentary on the pseudo-Aristotelian De plantis , Albert indicated that he
intended to treat plant life as a whole, discussing only those features which are
common to all plants, because “as Plato well said, the particulars [of plants]
are infinite and there can be no science of them.” 22 The same can be said of the

20) “... quod est compositius in rebus naturalibus tarn in corpore quam in anima. In corpore
enim habet elementa et commixtionem ipsorum et coagulationem et complexionem et compo-
sitionem ex etherogeneis et comunicationem cum anima vite manifeste, que est anima perfecta.
Et hanc compositionem aut non habent aut non tantam habent alie res naturales; simplicia enim
non habent earn, mineralia autem et plante non tantam habent ut animalia compositionem. In
animabus etiam animalium est sensus et operatio vite manifeste, que continent in se vegetatio-
nem et operationem vite occulte, sicut tetragonum continet in se trigonum. Propter utrumque
igitur istorum ultimam in rebus phisicis oportet esse scienciam animalium.” De animalibus I,
tr. 1, c. 1 (ed. F. Pelster, “Die ersten beiden Kapitel der Erklarung Alberts des Grossen zu De
animalibus in ihrer urspriinglichen Fassung,” Scholastik 10 [1935] 233.3-14). See also Physica I,
tr. 1, c. 4 (ed. Colon. 4/1:7.59-64).

21) “Quia vero in omnibus his [operationibus] que sciuntur de animalibus omnia communicant
animalia, licet differant in modis principiorum, ideo unam scientiam oportet esse de tota anima¬
lium diversitate tarn secundum genera quam secundum species ipsorum.” De animalibus I, tr. 1,
c. 1 (ed.F. Pelster 233.15-18).

22) “... eo quod particulars [plantarum] sunt infinita, nec eorum sicut Plato bene dicit, potest
fieri disciplina.” De vegetabilibus I, tr. 1, c. 1 (ed. Ernst Meyer and Carl Jessen [Berlin: Georg

M. W. Tkacz / Vivarium 45 (2007) 30-68

science of animals, for although individual animals manifest to sense percep¬
tion various individual features, that by which they are animals is the same in
all. This is why there is a single identifiable science of animals, says Albert,
“because their principles are the same in each, whether they are principles of
their reproduction, life, nutrition, or behavior.” 23 Moreover, this is true even if
various species of animals differ in the way in which they participate in these
common principles. 24 It is clear, then, that the zoologist should not discuss
every animal singly, because such a discussion would be repetitive, returning
to the same principles again and again. Rather, the task of the researcher is to
investigate those features common to animals and distinguishing them as dis¬
tinct forms of life as well as those features common within animal species and
distinguishing one species from another. 25

Nonetheless, common principles are studied for the sake of understanding
individuals in terms of the principles. Thus, the value of articulating principles
in the study of nature is to allow research into the various species of natural
subjects in which the principles are exemplified. 26 Albert makes this point
directly in addressing the objective of scientific research:

In natural science, however, it is necessary not only to consider change universally accord¬
ing to a common description, but it is necessary to focus so that primary change in indi¬
viduals, especially in sensible animals, may be determined. This is because in natural science
we must investigate universal principles through individuals, for in such a science particu¬
lars are more known than universals. Through individuals we hold that it is consistent and
necessary that universals and principles of universals exist, because we accept those univer-

Reimer, 1867] 4.6); see Karen Reeds, “Albert on the Natural Philosophy of Plant Life,” in Alber-
tus Magnus and the Sciences, 341-54, esp. 344-45.

23) “Quia enim principia ipsorum eadem sunt sive sint generationis sive vite sive nutrimenti sive
regiminis vite ipsorum .. De animalibus I, tr. 1, c. 1 (ed. F. Pelster 233.18-20).

24) “... licet differant in modis participandi principia ilia, sicut iam diximus.” De animalibus I,
tr. 1, c. 1 (ed. F. Pelster 233.20-21).

25) “lam autem diximus in praehabitis quod si loquamur de quolibet sigilladm animali, sermo
talis erit causa redeundi in eumdem sermonem multotiens, eo quod multis res communiter acci-
dentes eaedem sunt in multis animalibus.” De animalibus XI, tr. 2, c. 3 (ed. Stadler 15:795.1-4).

26) “... licet una sit communis scientia physicorum, quae est de mobili corpore in communi:
haec enim non sufficit, nisi habeantur etiam scientiae propriae de rebus propriis et appropriatis
in natura.” De animalibus XI, tr. 1, c. 1 (ed. Stadler 15:762.33-36). Note also Alberts warning
against a Platonic conception of such principles that separates them from their material exempli¬
fication in De animalibus XI, tr. 2, c. 4 (ed. Stadler 15:797.15-18).

M. W. Tkacz / Vivarium 45 (2007) 30-68

sals that are exemplified in particulars and we reject those universal not exemplified in
particulars.27

The zoological researcher is concerned first of all with the morphology and
behavior common to all species of animals and, then, getting down to detailed
cases, studies the morphology and behavior of specific species. 28

Sciences are distinguished by their subjects and the definition of a self-
contained subject domain makes possible research into the common proper¬
ties of the substances composing the domain. Albert recognizes that Aristotle s
libri naturales mark off a distinct domain of living (ensouled) substances hav¬
ing organic properties existing for the sake of realizing their capacities for
various life functions. This domain is divided according to the way in which
these life-functions exist and operate in the living substance. Those living
things that exhibit only those basic life functions of nutrition, growth, and
reproduction are distinguished from those that exhibit, in addition to these
functions, sensation and locomotion as well. This last group provides the sub¬
ject domain for zoological research. These divisions into subject domains for
study are possible because an analysis of sense perception in terms of “existing
for the sake of” reveals the existence of substances exhibiting these various life
functions. This allows research to proceed in its task of discovering the univer¬
sal principles common to living things in general as well as those common
within the various species of living things. Such a research program uncovers
these principles as they are exemplified in the individual substances to which
the researcher has access through sense perception. Zoology, therefore, can be
demarcated as a distinct science on the basis of its distinct subject domain — a
subject domain Albert finds clearly delineated in Aristotle’s biological works.

27) “In physicis autem oportet non solum accipere universale movens secundum rationem com-
munem, sed oportet descendere, ut videatur movens primum in singulis et praecipue in his quae
sensibilia sunt animalia, quia in physicis per singularia inquirimus rationes universales, eo quod in
physicis particularia digniora sunt universalibus, quia per singularia opinamur [esse] convenientes
et necessarias esse rationes universalium et universales, quia eas quae conveniunt particularibus,
acceptamus et eas quae non conveniunt particularibus, abicimus.” Deprincipiis motus processivi ,
tr. 1, c. 2 (ed. Colon. 12:49.20-31). See also Deanimalibus XI, tr. 1, c. 1 (ed. Stadler 15:764.3-14).

28) De animalibusYl, tr. 2, c. 3 (ed. Stadler 15:795.5-796.11). Note especially Albert s comment
that “prius debemus dicere operationes communes omnibus generibus animalium, et deinde
descendendo dicemus operationes attributas cuiuslibet formae sive speciei animalium.” (ed.
Stadler 15:795.15-18.). See Benedict M. Ashley, “St. Albert and the Nature of Natural Science,”
in Albertus Magnus and the Sciences , 73-102, esp. 87-94.

M. W Tkacz / Vivarium 45 (2007) 30-68

III. Hie End of Zoological Research

In one of the earliest texts of his Aristotelian commentaries, Albert presents an
argument against those who would deny the possibility of natural science. 29
He considers a series of three objections to the scientific knowability of natural
subjects, ascribing these objections to Heraclitus. The first is that the proposed
subjects of natural science, natural forms, exist in an infinite variety and so
cannot be understood by the finite human intellect. 30 The second objection is
that the natural scientist is unable to construct definitions of natural subjects
that can serve as the middle terms of scientific demonstrations, for such defi¬
nitions will apply to natural individuals equivocally. 31 The changeability of
natural forms is the basis for the third objection that such instability and con¬
stant motion prevents natural forms from being the subject of scientific dem¬
onstration which always concerns the stable and the necessary. 32

Albert replies to the first two of these objections by clarifying the precise
subject of natural science in terms of the goal of research. With respect to the
first objection, he points out that the aim of nature is to produce an ens com -
pletum as the perfection of an individual of a kind. This completed being is
finite and it is so because it is produced by its essential causes of form and
matter received through the agency of a moving cause and the end toward
which the agency is tending. 33 The subjects of scientific research are not indi¬
viduals in their infinite variety, but the causes of entia completa that produce
them as entia of this or that kind and the variety of kinds is finite. In a similar

29) Physica I, tr. 1, c. 2 (ed. Colon. 4/1:3-5). For a summary of the text and its relation to Alberts
account of scientific explanation see William A. Wallace, “Albertus Magnus on Suppositional
Necessity in the Natural Sciences,” in Albertus Magnus and the Sciences , 103-28, esp. 111-14.

30) “Sed physica secundum suum esse et in eo quod sunt et secundum quod sunt, sunt infinitas
differentias habentia. Intellectus autem omnis refugit infinitum et nullo modo comprehendit
ipsum. Naturalia ergo vere non comprehenduntur.” Physica I, tr. 1, c. 2 (ed. Colon. 4/1:3.50-55).

31) “Adhuc autem, secunda ratione negat Heraclitus physicum negotium posse habere medium,
per quod aliquid probari possit et demonstrari, quia diffinitio tale medium erit, si medium habe-
bit. Sed diffinitio in physicis est particulars, ut dicit, quoniam si accipiatur diffinitio animalis
secundum esse physicum, ipsa secundum unumquodque animal erit altera et secundum esse
cuiuslibet particularis erit altera. Sicut ergo non est unum esse plurium, ita nec una diffinitio nisi
aequivoce.” Physica I, tr. 1, c. 2 (ed. Colon. 4/1:3.56-65).

32) “Tertia Heracliti ratio est formarum naturalium instabilitas; accepta enim forma secundum
esse, quod habet in materia, mutatur, eo quod omnia physica motui subiacent et mutationi;
numquam ergo in eodem statu permanent. De talibus autem non est scientia, quae solummodo
de necessariis est.” Physica I, tr. 1, c. 2 (ed. Colon. 4/1:4.1-6).

33) Physica I, tr. 1, c. 2 (ed. Colon. 4/1:4.42-60); note esp. (ed. Colon. 4/1:4.55-60): “Ens enim
completum intenditur a natura, et hoc finitum est et finitur per causas suas essentiales realiter

M. W. Tkacz / Vivarium 45 (2007) 30-68

way, Albert argues, Heraclitus’ second objection overlooks the fact that the
definitions of the natural scientist apply to what is true of individuals as spe¬
cies and not to individuals as such. Univocal definitions covering many indi¬
vidual cases are possible because, even though a great diversity of individual
differences arise from the dispositions of matter, such differences are never the
end at which nature aims. 34

Regarding the third objection, Albert notes that Heraclitus’ position is
essentially that of Ptolemy who argued that the diversity of opinion among
natural scientists bears witness to the instability and non-necessity of natural
forms. Thus, Ptolemy was convinced that there can never be a true science of
natural subjects, as there is of mathematical subjects, but only opinion. 35
Albert realized that Ptolemy’s position was not based simply on this analogy,
for in his commentary on the Posterior Analytics Albert explains that

Ptolemy says in the first book of the Almagest that man ought not to satisfy his mind with
probabilities and opinions, because these do not bring about stable concepts in the mind,
but only be satisified with demonstrated and certain things which certify and establish
understanding because they are certain and eternally stable. 36

Because scientific explanation is always a matter of demonstrating a necessity
in the subject and natural subjects are radically contingent, there can be no
true science of them. This is why the comparison with mathematics, the sub¬
jects of which are absolutely necessary, provides for Ptolemy a telling contrast.

acceptas, quae sunt forma et materia, et per causam moventem, quae est efficiens, et per causam,
ad quam est motus, quae est finis.” Cf. Albert’s later approving citation of Plato on this point in
his De vegetabilibus\ see note 22 above.

34) Physics I, tr. 1., c. 2 (ed. Colon. 4/1:4.61-80); note esp. (ed. Colon. 4/1:4.74-80): "... quo-
niam in speciebus per principia essentialia diversificatur, sed in individuis multiplicatur tantum
per dispositiones materiae, ad quae numquam ex principali intentione respicit natura.”

35) “... etiam vir in multis prudens, Ptolemaeus, propter ultimam rationem [Heracliti] dicit de
naturis non haberi scientiam certam propter sui mutabilitatem, sed potius esse opinionem de
ipsis, cuius signum esse dicit, quia plurimi in naturis diversa opinati sunt, omnes autem in mathe-
maticis consone tradiderunt.” Physica I, tr. 1, c. 2 (ed. Colon. 4/1:4.26-32).

36) “... dicit Ptolemaeus in primo Almagesti, quod non probabilibus et opinabilibus debet homo
replere animam suam, quia non faciunt stantem habitum in anima, sed demonstrativis et certis
quae certificant et stabiliunt intellectum, quia certa sunt et aeternaliter stantia.” Posteriora Ana-
lytica I, tr. 1, c. 1 (ed. Borgnet 2:2b). Here, as in other texts, Albert cites Ptolemy with approval in
contrast to the discussion in the Physica (cited in note 35 above); on the apparent inconsistency see
William A. Wallace, “The Scientific Methodology of St. Albert the Great,” in Albertus Magnus Doc¬
tor Universalis 1280/1980 , ed. Gerbert Meyer and Albert Zimmermann (Mainz, 1980), 385-407.

M. W Tkacz / Vivarium 45 (2007) 30-68

Albert agrees that scientific explanation is a matter of demonstrating neces¬
sity in the subject, yet he does not accept Ptolemy s claim that natural subjects
lack the requisite necessity:

We, however, say that there is a science and demonstration of natural things, because natu¬
ral things have a subject, attributes, and principles through which an attribute is proved of
a subject; otherwise there could be no understanding or knowledge of these things which
concern all the sciences and which Aristotle, nevertheless, indicates do exist in nature. 37

In response to the third objection, then, Albert is able to point out that just as
the essential species of the natural subject allows univocal reference, so abstrac¬
tion from individuating matter allows a demonstration of necessity in the
natural subject. The contingent and unstable can be the subject of scientific
research because such research seeks to explain it in terms of its being the kind
of thing that is capable of change, rather than in terms of the change itself. 38

Albert s digression on the possibility of natural science is a defense of theo¬
retical knowledge of natural subjects in the Aristotelian sense outlined in the
Posterior Analytics . 39 Research aimed at attaining such knowledge does so by
establishing univocal definitions of subjects considered as species, by discover¬
ing their causes, and by providing causal demonstrations of those accidents
that belong to the subject by necessity.

In his first commentary on Aristotle’s De partibus animalium , Albert opens
his consideration by raising two questions: whether science is a dual process
and whether description is more necessary to science than the assignment of
causes. 40 After considering a series of objections, Albert affirms that science
involves two processes. One process is description ( narratio ), which he associ-

37) “Nos autem de physicis dicimus esse scientiam et demonstrationem, cum habeant physica
subiectum et passiones et principia, per quae passio de subiecto probatur; aliter enim non esset
intelligere et scire in ipsa, quod est circa omnes scientias, quod tamen innuit Aristoteles esse in
physica.” Physica I, tr. 1, c. 2 (ed. Colon. 4/1:4.36-41).

38) “Per hoc patet solutio ad ultimum, quia sicut quiditas et esse univocantur, ita etiam immobi-
lia sunt et necessaria, non quidem immobilia sicut per diffinitionem a motu et ratione motus
separata, sed sicut ratio rei mobilis est immobilis, eo quod ipsa abstracta est a particulari, quod
secundum actum est in motu et mutatione.” Physica I, tr. 1, c. 2 (ed. Colon. 4/1:4.81-5.1).

39) See Physica I, tr. 1, c. 5 (ed. Colon. 4/1:8.28-31): “Dico autem, quod omnis scientia, quae
habet principia, sic procedit, et ilia sola est vere scientia, quia est demonstrativa et effectus solius
demonstrationis est scire.” Note the reference to the Posterior Analytics at ed. Colon. 4/1:8.47-51.

40) “Utrum in scientia sit modus processivus duplex” and “Utrum processus narrativus sit
magis necessarius quam causarum assignativus.” Quaestiones de animalibus XI, q. 1-2 (ed. Colon.
12:218-19).

M. W. Tkacz / Vivarium 45 (2007) 30-68

ates with teaching (< docens ). Another is the assignment of causes (assignatio
causarum ), which he associates with learning (discern). 41 Yet, the process of
learning or scientific research also involves description, for the causes of a
subject cannot be known in the absence of a descriptive account of existence
and attributes of the subject. 42 Description, therefore, is for the sake of either
teaching or learning. The assignment of causes, however, is for the sake of
both, because the assignment of causes is the process that yields knowledge,
the end of both teaching and learning. From these considerations, Albert
draws the conclusion that, while description is necessary to both teaching and
learning, causal explanation is “more necessary.” 43

As is the case with any Aristotelian science, zoology can be considered in
two ways. First, it can be considered as a doctrine (scientia docens ), an articula¬
tion and systematic exposition of what has been established by demonstration
of its subject domain. Albert tends to focus on this sense of scientia in his
Posteriora Analytica , where he treats the nature of science in terms of a general
logic of demonstration. 44 Second, zoology can also be considered as an inves¬
tigative enterprise, a research program directed at attaining demonstrative
knowledge of its subject domain. This notion of science as dialectica inquisi-
tiva is the characterization most commonly found in Albert s logical works as

4,) “Dicendum, quod duplex est processus in scientia. Et hoc manifestum est ex parte rei et ex
parte nostra. Ex parte rei, quia in qualibet scientia supponitur aliquid tamquam fundamentum
scientiae, et ex isto in sequentibus [vel] consequentium causae assignantur; sed suppositiones
non traduntur nisi narrando, sed conclusiones consequentes traduntur causas assignando; ergo
[in scientia duplex est processus]. Ex parte nostra requiritur uterque processus, quia ad docentem
pertinet narrare et ad discentem, cum dubitat, causas quaerere.” Quaestiones de animalibus XI,
q. 1 (ed. Colon. 12:218.36-45).

42) “... sed causae cognitio non potest esse sine aliqua narratione, quia non potest induci sine
suppositione principii...” Quaestiones de animalibus XI, q. 2 (ed. Colon. 12:219.23-25).

43) “Ad istud dicendum, quod uno modo primus processus est magis necessarius et alio modo
alius. Audientibus enim provectis necessarior est processus secundus, sed minus provectis neces-
sarior est processus narrativus.” Quaestiones de animalibus XI, q. 2 (ed. Colon. 12:219.11-15).
See also Quaestiones de animalibus XI, q. 2 (ed. Colon. 12:219.6-10): “[Philosophus] dicit enim,
quod processus narrativus est propter docentem vel ex parte docentis, sed processus causarum
assignativus est propter utrumque; ergo secundus est magis necessarius.”

44) “In scientia autem syllogistica ultimum et optimum ... est demonstratio, ideo est finis pro-
ximi operis tradentis notitiam syllogisticam. Sed quia omne instrumentum et organum ad aliq¬
uid aliud refertur, ideo finis ultimus intendentis de syllogismo est scientia demonstrativa, hoc est
scientia docens modum et artem demonstrandi universaliter in qualibet particulari scientia
demonstrativa, sicut sunt scientiae mathematicae, geometria, astronomia, et musica, et multa
alia.” Posteriora Analytica I, tr. 1, c. 1 (ed. Borgnet 2:la-b).

M. W Tkacz / Vivarium 45 (2007) 30-68

well as his libri naturales . A5 It is especially in this second sense of science that
zoology is a dual-phased activity of description and causal demonstration. The
descriptive phase provides an account of the subject domain, animal life,
through rigorous consideration of common notions, review of previous
research, new observation and measurement, dialectical organization of the
data through careful definition and taxonimic arrangement, and so on. All of
this is carried out for the sake of eventually demonstrating the properties so
discovered as necessary of the subject domain. Albert makes it quite clear that
this is precisely what he finds in Aristotle’s libri de animalibus : an example and
recommendation of a research program aimed at theoretical knowledge of its
subject through causal explanation. 46

IV. The Principles of Zoological Research

In commenting on Aristotle’s distinction between true unqualified and sophis¬
tical knowing in the Posterior Analytics, Albert remarks that we can have true
unqualified knowledge only when we understand the cause of the subject
under study. He then adds:

We say “when we understand the cause? not causes in the plural, because there is one origi¬
nal cause, the causes of causes, which is the end. It is through this cause above all that
something is known, because each thing is determined and known with respect to its max¬
imal end and its essential optimal state. Cause is said in such a way that it is understood as
one with respect to the subject, for there are three things which together constitute the
principle of knowing: the agent, the form, and the end, as Aristotle explains in the second
book of the Physics. Moreover, although there are many causes of any given thing, one is
always the completing cause, which is the cause above all, and it is with respect to that
completing cause that it is said: knowing is when we understand the cause. 47

45) Alberts most detailed treatment of zoology as dialectica inquisitiva is in book XI of his De
animalibus*. tr. 1, c. 1 distinguishes exposition of the science of animals as an act of teaching from
research on animals based on observation (ed. Stadler 15:762.22-38 and 763.6-24); tr. 1, c. 2
discusses the two stages of animal research (ed. Stadler 15:765.3-27); tr. 2, c. 3 remarks on the
order of zoological investigation (ed. Stadler 15:793.13-29 and 795.1-34). For Alberts general
treatment of dialectica inquisitiva , see William A. Wallace, “Albert the Greats Inventive Logic: His
Exposition of the Topics of Aristotle,” American Catholic Philosophical Quarterly 70 (1996), 11-39.

46) See, for example, the explicit statement at Quaestiones de animalibus XI, q. 1 (ed. Colon.
12:218.45-50) discussed in more detail in section V below.

47) “Cum autem dicimus cum causam cognoscimus , non causas in plurali: ideo quia una est princi¬
palis causa, quae est causa causarum, quae est finis, per quam potissime scitur; quia unumquodque
maxime fine suo et optimo essentiali determinatur et scitur. Dicitur etiam causa ut subjecto intel

M. W Tkacz / Vivarium 45 (2007) 30-68

Albert here identifies a fundamental principle of investigation in any Aristote¬
lian research program. While there are many explanatory factors making pos¬
sible understanding of a subject, each of these factors contributes to
understanding only in virtue of the factor indicating that for the sake of which
all the others are operative in the subject. Confronted with a material subject
to be explained, the researcher seeks to discover the causes productive of the
subject s being and operation. Such causes will involve those efficient agencies
operating upon the material as well as the material morphologically instanti¬
ated in the subject. Yet these explanatory factors can produce understanding
in the researcher only when seen in light of the subjects maximal state of
development which is its essentially optimal state of being. In other words, the
researcher comes to understand why the subject is the way it is, not simply by
knowing that of which it is composed, how its compositional elements are
morphologically arranged, and what agencies brought about that arrange¬
ment, but how all of these factors are related to the subject considered as an
ens completum. It is the subject considered as a completed being of a certain
kind, then, that is the object of research activity. The causes of the subject so
considered constitute its explanation only with respect to its completed being,
or as traditionally termed, final cause . 48

Albert contrasts scientific explanation understood in this way to the con¬
ception of the early Greek natural philosophers who, limiting explanation to
material necessity, denied final cause:

All the ancient physicists introduced that cause which is assumed to be necessary on account
of the material, and they neglected final cause, saying that nature does nothing on account
of a determined end. Rather, this or that is thought to act from the necessity of the material.
This or that thing, such as men or asses or long or short things, exists from necessity and
comes to be from necessity, and they are born apt to become so on account of their material
dispositions, which compel by inclining more to this effect rather than another, according
to how hot or cold the material is. 49

ligatur una: quia tres quae principium sunt sciendi, in unam coincidunt, efficiens et forma et finis,
ut dicit Aristoteles in secundo Physicorum. Ad hue autem quamvis multae sint causae alicujus, una
autem semper est completiva quae potissime est causa, et ad illam respiciendo dicitur, quod scire
est cum causam cognoscimus.” Posteriora Analytica I, tr. 2, c. 1 (ed. Borgnet 2:23a).

48) Albert discusses the application of final cause explanation in natural philosophy in his Physica II,
tr. 2, c. 4 on the kinds of final cause, c. 5 on the ordering of causes to final cause, c. 22 on the
four causes and “that for the sake of which” in nature, and tr. 3, c. 4-6 on explanation given the
supposition of the end (ed. Colon. 4/1:104-06, 130-32, 139-45).

49) “Omnes autem antiqui physici inducunt se in causam istam quae est necessitas ex parte
materiae sumpta, et negant causam finalem dicentes, quod natura nihil agit propter finem

M. W. Tkacz / Vivarium 45 (2007) 30-68

The ancient naturalists were mislead by their neglect of that for the sake of
which in explanation to introduce chance as a cause of natural subjects. Fol¬
lowing Aristotle, Albert picks out Empedocles as an apt example. The parts of
animals are arranged, according to Empedocles, not for the sake of the utility
intended by nature, but by chance alone. The development, morphology, and
arrangement of animal teeth, for instance, are not explained by the functional
requirements of the animal, but come to be by chance out of the material of
which they are composed. On this view, the explanation of the presence and
shape of molars in the mouth of a certain species of animals makes no refer¬
ence to the food-grinding function of molars. Rather, the explanation simply
indicates the material of which molars are composed, along with its tendency
to be hard and hold a shape, and that this material comes to be molar-shaped
and positioned at the back of the mouth by chance . 50

Albert replies that he disagrees with those who completely neglect the role
of chance in nature just as much as he disagrees with those who claim that all
natural things come to be by chance . 51 Nonetheless, he points out that the
reason why it is impossible that any natural thing comes to exist by chance
alone is that it would not thereby be natural. What is natural is, using the
traditional Aristotelian formula, what exists always or for the most part {sem¬
per vel frequenter ); in contrast, chance events are those that, by definition,
occur by accident and luck {a casu etfortuna ). 52 While nature can be identified
with either the material components of the subject or its form, it is the form
that determines the way the material will exist and function in the subject. It

determinatum, sed cogitur hoc vel illud facere ex necessitate materiae, dicentes, quod hoc vel hoc
ex necessitate sunt et Hunt ex necessitate et fieri apta nata sunt, scilicet homines vel asini vel longa
vel brevia, propter materiales dispositiones, quae cogunt inclinando ad hunc effectum plus quam
ad alium, sicut quod calor talis et frigus huiusmodi est materia.” Physica II, tr. 3, c. 1 (ed. Colon.
4/1:132.67-77).

50) “Quaerunt autem dicti philosophi dicentes, quod sicut in transmutatione elementorum
ea quae fiunt, fiunt ex necessitate, quid prohibet quin ea quae fiunt in animalibus, similiter fiant
ex necessitate materiae, sicut dentes anteriores fiunt tenues in acumine et acuti, ita quod cibum
faciliter dividunt, et maxillares dentes, qui et molares dicuntur, fiunt lati utiles ad conterendum
et molendum cibum, non propter hoc quod natura has utilitates intendat in forma dentium,
sed quia talis erat materia in dispositione sua diversa . . Physica II, tr. 3, c. 1 (ed. Colon.
4/1:133.33-42).

51) .. quia sicut condemnamus eos qui casum negant esse in natura, ita condemnamus etiam
eos qui omnia naturalia dicunt fieri a casu .. ” Physica II, tr. 3, c. 2 (ed. Colon. 4/1:134.9-11).

52) “Sed ea quae sunt de numero eorum quae fiunt a casu et fortuna, nullum fit semper vel fre¬
quenter ... ergo ea quae fiunt a natura, non fiunt a casu et fortuna.” Physica II, tr. 3, c. 2 (ed.
Colon. 4/1:134.14-17).

M. W Tkacz / Vivarium 45 (2007) 30-68

is form, then, that is nature in the primary sense and it is so in virtue of its
orientation to function . 33 Resisting the Empedoclean identification of chance
and necessity, Albert admits that explanation does require the uncovering of a
necessity, but he insists that natural necessity can only be with respect to the
end . 54 Appealing to chance, therefore, can never provide the appropriate sort
of necessity. Indeed, if Empedocles is right, nature cannot be understood at
all, even though there is a kind of determination by the material. To claim that
a natural subject exists and functions by chance is to say that it cannot be
explained . 55

Returning to Empedocles’ example of animal teeth, Albert s account of the
Aristotelian explanation provides a characteristic contrast . 56 Whatever chance
events may be involved in the formation of teeth, such events cannot be the
explanation of animal teeth. Certain animals have teeth by nature—that is,
they are regularly born with teeth and being toothed in a particular way is a
morphological indication of their species. Even if such a morphology did
come to exist by chance in a particular instance, once it becomes characteristic
of the species it is no longer chance. Thus, its explanation cannot simply refer
to that chance event. Following Aristotle, Albert locates the true explanation
in the function that makes the morphology intelligible as a regular and defin¬
ing characteristic of the species. Dogs, for example, have sharp teeth because
they are carnivors and must tear their food. The dental morphology of dogs,
then, is explained by their carnivorous nature . 57 Further, the identification of
finality as an explanatory principle of zoological research allows for the func¬
tional explanation of analogous forms:

53) “Quia vero natura dupliciter est et dicitur, alia enim est materia, alia forma, finis autem in
naturis forma est, sicut constat per ante dicta, erit forma causa finalis, cuius gratia sunt et fiunt,
quaecumque sunt et fiunt in naturalibus ante formae adeptionem.” Physica II, tr. 3, c. 2 (ed.
Colon. 4/1:136.3-8). See also Alberts examples from plant and animal life in the preceding pas¬
sage (ed. Colon. 4/1:135.45-136.3).

54) “Si igitur omnia naturalia aut a casu videntur fieri aut a natura propter aliquid operante et
finem intendente et impossibile est ea fieri a casu et fortuna, necessario fiunt a natura determi-
natum finem in unoquoque eorum intendente.” Physica II, tr. 3, c. 2 (ed. Colon. 4/1:134.27-31).

55) To his reply to Empedocles Albert adds two further points: in Physics II, tr. 3, c. 3 (ed. Colon.
4/1:136-39) he discusses the traditional Aristotelian analogy of nature and art and in Physics II,
tr. 3, c. 4 (ed. Colon. 4/1:139-40) he argues that acting for an end does not necessitate conscious
intention of the end.

56) De animalibusX. II, tr. 3, c. 6 (ed. Stadler 15:883-888).

57) The example of dog teeth is given by Ashley, “St. Albert and the Nature of Natural Science,”
80. For general remarks on the teeth of carnivores, see De animalibus XII, tr. 3, c. 6 (ed. Stadler
15:883-84).

M. W. Tkacz / Vivarium 45 (2007) 30-68

As walking animals of prey have sharp teeth, so birds of prey have curved beaks and sharp
curved talons so that they are able to seize and hold their prey in their talons while flying
and tear at what they are holding with their beak. 58

The morphological identification, description, and definition of animal species
depends on grasping that for the sake of which the form exists. Consequently,
the zoological researcher attains the explanation of the morphology when that
for the sake of which the form exists is demonstrated . 59

As already noted, Albert’s rejection of Empedocles’ identification of chance
and necessity does not imply that Albert held scientific research to be anything
other than the progressive disclosure of necessity. Rather, the importance of
that for the sake of which in explanation indicates that the necessity uncov¬
ered by the natural scientist is not the necessity of force that Empedocles had
in mind. The error of the early Greek naturalists, then, provides Albert with
the occasion for a detailed discussion of the type of necessity found in animals,
plants and other natural things . 60

Commenting on the entry on the modes of necessity in Aristotle’s philo¬
sophical lexicon , 61 Albert points out that necessity is not limited to absolute
compulsion (necessitas absoluta). Necessity may also be suppositional ( necessi -
tas suppositionis ) when what is prior is inferred from what is posterior. While
absolute necessity is associated with material cause, suppositional necessity
concerns final cause. Given that the putrid humors are to be driven from the
body, it is suppositionally necessary that the medicine be drunk. Necessity is

58) “Sicut autem in animalibus gressibilibus rapacibus sunt dentes acuti, ita in avibus rapacibus
sunt rostra curva et ungues acuti et curvi, ut volando rapere et retinere possit ungue et rostro
dilacerare quod tenuerit.” De animalibus XII, tr. 3, c. 6 (ed. Stadler 15:888.14-17).

59) The zoological investigator is also interested in how the material elements of which animals are
composed are present for the sake of the morphology. See Albert’s distinction between the finis
generations and the fines reigeneratae in De animalibus XI, tr. 2, c. 4 (ed. Stadler 15:797.15-29)
discussed in section V below.

60) “Causa autem omnis erroris istius quo antiqui inducti sunt ad credendum, quod ex necessi¬
tate materiae fieret omnis generatio animalium casu et plantarum et aliarum rerum naturalium,
fuit necessarium in disciplinis demonstrativis, in quo videbant, quod ex necessitate primorum
principiorum sequitur esse consequens, quia conclusio ibi necessario fit propter principiorum
materialium necessitatem, quae principia materialia sunt propositiones, inquantum in ipsis est con¬
clusio posita, et per habitum scientiae confiisum, sicut forma est in materia, sicut superius diximus.
Et ideo oportet nos investigate de isto necessario, qualiter est in physicis et qualiter in demonstrati¬
vis, et ostendere dissimilitudinem inter ilia.” Physica II, tr. 3, c. 5 (ed. Colon. 4/1:140.76-141.10).
6,) Metaphysica V, tr. 1, c. 6 (ed. Colon. 16/1:220-22); Albert provided an earlier account of the modes
of necessity based on the same text in his In I Sententiarum, dist. 6, A, art. 2 (ed. Borgnet 25:198a-b)
where he distinguishes necessitas simplex and necessitas in respectu finis.

M. W Tkacz / Vivarium 45 (2007) 30-68

the same, says Albert, in mechanics and in nature. If there are to be soldiers,
then there must be the fabrication of arms and, if humans are to exist, then a
certain complex of humors is necessary . 62 Thus, necessity is not limited to
absolute compulsion. What is necessary can also be that required on the sup¬
position that something is the case. Because final cause is a principle of zoo¬
logical inquiry, the sort of necessity disclosed in zoological explanations will be
of this suppositional kind.

Albert confirms this point in his discussion of natural necessity in his com¬
mentary on the Physics :

We ask first, therefore, whether the necessity of physical things is absolutely necessary or
necessary on the supposition and condition of some presubscribed presupposition. For
example, there is absolute necessity, as the necessity of the heavy thing descending and the
light thing ascending; it is not required that something be presupposed for this to be neces¬
sary. There is also necessity from a condition, the necessity which requires that something
be presupposed, nor is it itself necessary except on a supposition. For example, it is neces¬
sary for you to sit, if I am to see you sitting. 63

Albert continues pointing out that absolute necessity is found only in the
aptitude and necessity of matter. What is necessary, however, can also be
according to what is supposed (secundum suppositionem) on the grounds of
some hypothesis (in ordine hypothesis alicuius). This is the case when it is neces¬
sity that one sleeps, if ones sensory powers are to be rested . 64 One need not
sleep, but in order that the condition be fulfilled, sleep is necessary . 65

62) The examples are given in Metaphysical ', tr. 1, c. 6 (ed. Colon. 16/1:221.10-31).

63) “Quaeremus ergo primo, utrum necessarium physicorum sit necessarium simpliciter vel sit
necessarium ex suppositione et condicione cuiusdam praesuppositi et praestituti. Verbi gratia
necessarium simpliciter est, sicut necesse est grave descendere et leve ascendere; non enim oportet
aliquid praesupponi ad hoc, quod sit hoc necessarium. Necessarium autem ex condicione est, ad
cuius necessitatem oportet aliquid praesupponi, neque ipsum est necessarium nisi ex suppositione, et
sic necesse est sedere te, si ego video te sedentem.” Physica II, tr. 3, c. 3 (ed. Colon. 4/1:141.11-20).
645 “Est autem necessarium simpliciter in sola materiae aptitudine et necessitate. Necessarium
autem secundum suppositionem est in ordine hypothesis alicuius, sicut necesse est te dormire, si
quiescere debeant in te virtutes sensibiles.” Physica II, tr. 3. c. 5 (ed. Colon. 4/1:141.20-25).

65) William A. Wallace has studied the various attempts to capture Aristotle’s meaning in these
texts on suppositional necessity in “Aristotle and Galileo: The Uses of Hypothesis (Suppositio) in
Scientific Reasoning,” in Studies in Aristotle, ed. Dominic J. O’Meara, Studies in Philosophy and
the History of Philosophy, vol. 9 (Washington, DC, 1981), 47-77. Wallace further argues that
Albert’s commentaries on the libri naturales clarify what many ancient and modern commentaries
leave obscure, in “Albertus Magnus on Suppositional Necessity in the Natural Sciences,” in
Albertus Magnus and the Sciences , 103-28.

M. W Tkacz / Vivarium 45 (2007) 30-68

Again referring to the ancients who held that all natural things come to be
by chance through some material force (materiae obligatio ), Albert argues
that this is

just as if someone should say that it is not on account of the supposition of the end, but on
account of the demands and aptitude of the material that a house comes into being, because
he thinks that the wall is made erect, not to support the roof, but because the wall is com¬
posed of various materials: some of them heavy so that they necessarily descend to the
foundation as they are naturally disposed to move; some of them moderately light so that
they adhere to the heavy material in the lower parts and extend to contact the higher parts.
Thus, the expanse of the wall comes to be: stones are carried to the bottom and form the
foundation, wooden parts of moderate lightness go up higher, and the lightest on top. So
they say that there is a motion of the component parts in the composite, and thus from the
necessity of the material the structures, shapes, and forms compatible with their motions
come to be in those parts. Therefore, it is clear that, according to them, form follows the
necessity of the material and not that the material is required on account of the form such
that the material is not sought by nature except on account of form. This is absolute neces¬
sity which exists from the binding force of the material. 66

Albert finds such an account of the existence and form of the wall radically
incomplete and lacking in explanatory power. Were such structures to arise
simply out of the “binding force of the matter” (ex materiae obligatione ), then
the function of the wall would be inexplicable as a function. To say that nature
is like this is to deny the intelligibility of nature, for there would then be no
true ens completum in terms of which matter and all natural agencies are to be
understood. The Heraclitean objection Albert had discussed earlier in his
Physica would then be telling, because an infinite variety of forms could be
produced out of the available matter and nature could never be known. Natu¬
ral necessity, however, does not operate the way the ancient naturalists claim:

66) “... sicut si aliquis dicat non propter suppositionem finis, sed propter materiae obligationes
et aptitudinem fieri domum, quia cogitat parietem esse factum erectum non ideo, quod sustineat
tectum, sed ideo quia componitur paries ex diversis, quorum quaedam sunt gravia, et ideo neces-
sario descendunt in fundamentum deorsum, eo quod sic moveri apta nata sunt; quaedam autem
sunt modo medio levia, et ideo ilia cohaerent gravibus in inferioribus et extenduntur altiora
contingendo, et ita fit expansio parietis; lapides enim deorsum feruntur et fiunt fundamentum,
et ligna medio modo levia ascendunt altius, et levissima sunt in summo. Ita dicunt, quod est
motus mixtorum in mixto, et ideo ex necessitate materiae adveniunt eis figurae et formae com-
petentes motibus eorum. Sic ergo patet, quod secundum istos forma sequitur necessitatem mate¬
riae et non propter formam requiritur materia, ita ut materia non quaeritur a natura nisi propter
formam. Et haec est necessitas absoluta, quae est ex materiae obligatione.” Physica II, tr. 3, c. 3
(ed. Colon. 4/1:141.28-47).

M. W. Tkacz / Vivarium 45 (2007) 30-68

But what they said is not true, for although form does not come about without the neces¬
sary material, nevertheless it does not come about because of the necessity of the material.
If it did, the material would not be required for the sake of the form, but the material would
be able to have any form which would follow from the necessity of its motions. In the same
way, the form of natural things would be subject to chance, because it would not be
intended by nature. But we say that the form does not exist because of this, unless we mean
to indicate the cause of necessity in the sense of what composes the subject, that is, the
material. Rather, as in artifacts, everything prior comes about on account of what is subse¬
quent, because things produced subsequently are ends and prior things are directed to the
end. Nevertheless, subsequent objects do not exist without prior things; just as it is in all
things, so it is in natural things. This is because the end in them is form, which is the prin¬
ciple of the whole process on account of which everything comes to be and exists. 67

If an artifact, such as a house, is to be used as an analogy for nature, then one
cannot simply attend to the materials and their properties. Sooner or later, the
purpose of using these materials in building the house must be given, other¬
wise the house, as house, cannot be understood—all that will be grasped will
be a list of building supplies along with their associated material properties of
hardness, flexibility, porousness, etc. So, Albert clarifies the analogy:

This is clear in the case of a house, because that which is the principle of all those things
which come to be in the house, is what the one who makes the house intends, and that is a
shelter from rains and storm and a storage for riches and a safe place for contents. On
account of that end, the material of the house is sought, prepared, and put together, and
everything which is made in the house, is made because of that. This is why we said above
that the end, which is first in knowledge, is the last in operation and being; it is the cause
of causes, because on account of it other causes cause what they cause. 68

67) “Sed dictum istorum non est verum, quia licet non sine materia necessaria fiat forma, non
tamen propter necessitatem materiae fit forma, quia sic non quaereretur materia propter formam,
sed posset habere materia formam, quaecumque sequeretur necessitatem motuum materiae, et sic
iterum forma naturalis supponeretur casui, cum non intenderetur a natura; et dicimus, quod forma
non fit propter hoc, nisi notetur causa necessitatis, quae est disponens eius subiectum, quod est
materia. Sed potius sicut in operibus artium omnia priora hunt propter posteriora, eo quod pos¬
terior facta sunt fines et priora sunt ordinata ad finem et tamen posteriora non sunt sine priori-
bus, ita per omnia est in rebus naturalibus, quia finis in eis est forma, quae totius operationis est
principium, propter quod fiunt universa et sunt.” Physica II, tr. 3, c. 3 (ed. Colon. 4/1:141.48-63).

68) “Et hoc quidem manifestum est in domo, quia id quod est principium omnium eorum quae
fiunt in domo, est, quod sibi proponit efficiens, qui facit domum, et hoc est operimentum a
pluviis et tempestate et contentivum esse divitiarum et salvativum esse contentorum. Et propter
ilium finem quaeritur materia domus et paratur et coniungitur, et omne quod fit circa domum,
fit propter illud. Et ideo diximus superius, quod finis, qui primus in cognitione est et ultimus in
operatione et esse, est causa causarum, quia propter ipsum causae aliae causant, quod causant.”
Physica II, tr. 3, c. 5 (ed. Colon. 4/1:141.63-74).

M. W Tkacz / Vivarium 45 (2007) 30-68

Reference to the materials and their absolutely necessary properties become
part of the explanation of the entire given natural form only insofar as they are
suppositionally necessary for the form to exist and function as observed. Sup¬
positional necessity, then, is a principle of nature in light of which material
necessities are understood. Natural forms exist for the sake of an end, but the
end is not to be found in the matter and its material efficencies. It is, as Albert
puts it, in ratione , and this reason provides the principle of the whole being
and operation of the natural entity being explained . 69

Fundamental to zoological research are two related principles. The first is
that animals are morphologically defined and classified by form considered
functionally rather than structurally. Eagles are birds of prey, not because they
have sharply curved pointed beaks and taloned feet, but they possess these
morphological features because they are birds of prey. The researcher may take
such features as evidence of the carnivorous nature of these birds, but this is
possible only because the researcher already knows what a carnivore is and
how the morphology makes it possible. Explanations of morphologies, then,
depend on understanding the function to which the morphology contributes
and cannot be limited to descriptions of material or agent causes. The second
principle is that animals are explained by the disclosure of that for the sake of
which a morphology exists through the demonstration of the necessity of the
morphology for that end. Necessity in animals is conditional and, therefore,
the task of the researcher is the uncovering of this necessity as respective to an
end. Absent such disclosure, the research must continue until the necessity of
the morphological elements becomes manifest in terms of their purpose with
respect to the life and well-being of the animal.

V. The Methodology of Zoological Research

In the course of discussing the stages of zoological research, Albert notes that
Aristotle makes an explicit methodological recommendation:

The Philosopher, wisest and most expert in the sciences, proceeds in this science [of animals]
by first describing and second investigating and assigning the causes of what is described,
showing and affirming that we ought to do the same. 70

69) “Finis enim non est in materia habens necessitatem, sed potius in ratione, quia sic est movens
artificem et efficitur principium totius operationis, et ideo sic ab ipso fluit et motus, quo movet
efficiens, et necessitas materiae, qua praeparatur ad finem consequendum ” Physica II, tr. 3, c. 3
(ed. Colon. 4/1:142.14-19).

70) “Et ideo Philosophus, tamquam sapientissimus et expertissimus in scientiis, in scientia
ista [de animalium] procedit primo narrando et secundo narratorum causas inquirendo et assig-

M. W. Tkacz / Vivarium 45 (2007) 30-68

As discussed above, the goal of zoological research is to produce a scientia
docens , an established body of knowledge concerning animal life. Achieving
this requires a set of methodologies that fall into two distinct stages of scien¬
tific investigation. First, the zoologist must develop a detailed and accurate
description of the subject of study. Having attained this, the zoologist can
proceed to determine and establish the causes of the subject. Together, these
two phases of research ultimately aim at demonstrated explanation of animal
morphologies and behaviors through their causes. Separately, each research
stage represents a complex of various dialectical procedures to be employed by
the zoologist in the effort to come progressively closer to causal explanation.
In both stages, some methods will be those that are used in any science whereas
others will be proper to the study of animals. In either case, the methodology
is determined by the nature of the subject and the requirements needed to
bring the investigator closer to the final end of the research activity.

A. The Order of Zoological Research

Albert finds this two-staged research methodology evident in the very arrange¬
ment of Aristotle’s zoological treatises. The editio vetus , produced by Michael
Scot from the Arabic , 71 provided the text from which Albert produced both his
quaestiones as well as his later paraphrastic commentary. Unlike the editio nova
translated directly from the Greek by William of Moerbeke , 72 this older version
combined the three longer libri de animalibus in a single work of nineteen
books. The first ten books correspond to the Historia animalium and these are
followed by the four books of the Departibus animalium and the five books of
the Degeneratione animalium. Albert structures his own De animalibus accord¬
ingly: the first nineteen books comprising his paraphrastic commentary on
Aristotle’s text and an additional seven books containing his own zoological
researches . 73 In the first ten books recounting Aristotle’s animal histories,

nando, ostendens, quod nos similiter debemus facere, vel annuens.” Quaestiones de animalibus XI,
q. 1 (ed. Colon. 12:218.45-50).

71) De animalibus: Michael Scot's Arabic-Latin Translation. Part Two: Books XI-XTV: Parts ofAnimab.
Part Three: BooksXV-XIX: Generation ofAnimab. Edited by Aafke M. I. van Oppenraaij. Aristoteles
Semitico-Latinus, vol. 5 (Leiden, 1992 and 1998). Part One: Books 1-X: History ofAnimab has not
yet been published.

72) De historia animalium translatio Guillelmi de Morbeka , ed. Pieter Beullens and Fernand Boss¬
ier, Aristoteles Latinus, vol. 17.2.1 (Leiden, 2000); De generatione animalium translatio Guillelmi
de Morbeka^ ed. H. J. Drossaart Lulofs, Aristoteles Latinus, vol. 17.2.V (Bruges-Paris, 1966).

73) Weisheipl, “Alberts Works on Natural Science,” 572-74 [13], see also 572 [11]. For a general
discussion of the contents and structure of Albert’s De animalibus , see the introduction of Kitchell
and Resnick in Albertus Magnus On Animab, 1-42, esp. 34-42.

M. W Tkacz / Vivarium 45 (2007) 30-68

Albert locates the descriptive studies that are presupposed by any attempt to
explain animal morphology or behavior. He distinguishes the content of these
books from that of the later books on the parts and generation of animals
which provide causal explanations . 74

Toward the beginning of his paraphrase, Albert recounts Aristotle’s general
theoretical description of the nature of likeness and difference in animal parts
and the distinction of the descriptive genera of parts, life histories, behavior, and
character. He then adds the important methodological statement:

All of these ways of being we have described here in a general way both with respect to their
agreement and differences. We will go on to sharpen the consideration of these things when
we investigate and consider them all in particular; he who wishes, then, to consider the
whole of what we say, will know that what we have said in these things concerning the
animal kingdom is true. We anticipated those things which we said concerning the general
sameness and difference of animals, so that the diversity of animals might be manifested
with all of their accidents. After this we will identify the causes of all of these. This is indeed
the natural and correct way to acquire knowledge and in this way what we wish to know
will be manifested with certainty. 75

This two-stage procedure of, first, describing the likenesses and differences
among animals in an effort to demonstrate their diversity and properties and,
second, demonstrating the causes of the diversity and properties is both natural
and proper. It is natural and necessary because any attempt to state the cause of
a subject requires some articulation of the subject. Indeed, the differences and
likenesses of animal parts must be grasped prior to demonstrating their causes,
for without some clear morphological description, it is impossible to under¬
stand what is to be causally explained. Thus, the correct method of investigat¬
ing animal morphology is to begin with an account or description before
proceeding to causal demonstration. Morphological explanation arises from

74) De animalibus XI, tr. 1, c. 2 (ed. Stadler 765.23-27) where the books on the parts and gen¬
eration of animals are introduced by contrast with the descriptive contents of the “preceding ten
books” of animal histories.

75) “Omnes autem hos modos narravimus hie sermone communi et convenientias et differentias
modorum istorum. Posterius autem subtiliabimus considerationem in hiis, quando investigabi-
mus et apprehendemus omnia haec in speciali: et qui voluerit tunc considerare totum hoc quod
dicemus sciet quod verus est sermo quern hie diximus in communitate animalium. Anticipavimus
enim ea quae diximus de communi convenientia et differentia animalium, ut manifestetur diver-
sitas animalium cum omnibus accidentibus quae accidunt eis. Posterius enim dicemus causas
omnium istorum. Haec enim via naturalis est et recta ad scientiam aquirendam et per illam erit certa
manifestatio eius, quod scire intendimus.” De animalibus I, tr. 2, c. 1 (ed. Stadler 15:38.34-39.9).

M. W. Tkacz / Vivarium 45 (2007) 30-68

morphological studies that provide accurate descriptions of animal parts in
the context of a general morphology of the animal possessing the parts. Once
such a description is attained by the zoologist with sufficient detail and accu¬
racy, then it becomes apparent precisely what requires causal explanation and
how such explanation must be given. The same is analogously true for the
explanation of animal behavior and function in general . 76

Given the way science is discussed in the Posterior Analytics, Albert realizes
that it is important that this dual method of description ( narrativus ) and
causal explanation ( causarum assignativus) be put in the context of Aristotle’s
theory of demonstration. He, therefore, carefully considers an objection to the
two-stage procedure. Because scientific explanation is knowing a subject through
its causes and causes are known when demonstrated, it seems that there is no
room for description in the process of coming to know a subject. A demonstra¬
tion, after all, is a syllogism producing knowledge ( syllogismus faciens scire) and,
therefore, scientific method is demonstrative and not descriptive . 77

Alberts reply makes an important association of the quia!propter quid distinc¬
tion of the Posterior Analytics with the narratiolassignatio causarum distinction.
Scientific research is, indeed, demonstrative in the sense that the goal of research
is the demonstration of the subjects causes. Yet, demonstration is itself twofold,
for there is demonstration of the fact ( demonstrate quia) and demonstration
of the reasons for the fact (< demonstratio propter quid). The latter is through the
cause ( per causam ), notes Albert, but the former is through the effect or from
a supposition ( per effectum aut ex supposition)™ The aim of the researcher is
to know the subject through the causes on account of which the subject is the
subject it is and possesses the properties it does. Before the investigator can
attain this knowledge, however, he first needs to know how the subject and its
properties exist. Moreover, this quia knowledge of the facts to be explained
cannot be the product of mere conjecture or guesswork, for then one could

76) Albert makes it clear that the two-staged method is not restricted to zoology, but is gener¬
ally required in any scientific research program. See, for example, De animalibusYl , tr. 1, c. 2
(ed. Stadler 15:765.15-23) where he finds the same general structure of research in both the
natural and mathematical sciences.

77) Quaestiones de animalibus XI, q. 1, obj. 1 (ed. Colon. 12:218.17-20): “Et videtur quod non
[scientia sit modus processivus duplex: narrativus et causarum assignativus]. Quia demonstratio
est syllogismus faciens scire. Cum igitur omnis scientia faciat scire, ergo omnis scientia erit
demonstrativa; nulla ergo erit narrativa.”

78) “Ad primam dicendum, quod duplex est demonstratio, scilicet propter quid et quia. Demon¬
stratio propter quid fit per causam, sed demonstratio quia habetur per effectum aut ex supposi-
tione.” Quaestiones de animalibus Y*. I, q. 1, ad 1 (ed. Colon. 12:218.51-55).

M. W Tkacz / Vivarium 45 (2007) 30-68

never be certain just what the propter quid demonstrations are supposed to
reveal about the subject. Rather, the facts about the subject must be estab¬
lished in some reliable manner and this requires a sort of demonstration. Such
quia demonstrations will concern the observed effects and arise out of the
efforts of the researcher to define his subject and its properties as well as to
distinguish these from other subjects with their distinctive properties. In doing
this, the zoologist is clearly considering the observed subject as an effect and
as presupposed with respect to his goal of causal explanation. While the assign¬
ment of causes ( assignatio causarum) is a matter of demonstrating propter quid ,
then, scientific description ( narratio ) also involves a demonstration: one that
establishes the facts to be causally explained. Therefore, both description and
demonstration are part of a zoological research program that has a generally
demonstrative character . 79

B. Zoological Description

As the purpose of the descriptive phase of zoological research is to establish
and articulate the morphological and functional types and properties of ani¬
mals, its goal concerns definition. More exactly, the aim of scientific descrip¬
tion is to provide, in some rigorous manner, definitions of animal species both
in terms of what distinguishes diverse kinds and what is common to them.
The specific means by which the zoologist attains the information used in the
articulation of the definition will differ from one research situation to another
depending on the investigators access to the subject of study. Thus, some
studies will focus on observations and measurements made in situ whereas
others may revolve around partly or wholly manipulated observational exper¬
iments . 80 Whatever the case, the dialectical means by which the researcher
proceeds from observation to definition will involve the method of division.

In a preface to his commentary on the Topics , Albert notes that the scientific
investigator requires a method by which he can reason about any problem that
arises concerning what is properly predicated of the subject of his investigation.
He goes on to explain that “by any problem” he is referring generically to any
question about inherence ( inesse ), such as inhering as an accident, or a genus,

79) Quaestiones de animalibus XI, q. 2 (ed. Colon. 12:218-19). On the distinction of quia and
propter quid demonstrations, see Posteriora Analytica I, tr. 3, c. 6 (ed. Borgnet 2:82a-84b).

80) Albert remarks that it is necessary to probe experience ( experimentum ) in more than one
way according to all circumstances ( secundum omnes circumstantias ) in Ethica VI, tr. 2, c. 23
(ed. Borgnet 7:443a). For a few zoological examples, see De animalibusV , tr. 1, c. 2 (ed. Stadler
13:415-16); VI, tr. 1. c. 6 (ed. Stadler 15:461); XV, tr. 1, c. 8 (ed. Stadler 16:1009).

M. W. Tkacz / Vivarium 45 (2007) 30-68

or a property, or a definition . 81 Moreover, inhering as a differentia is reducible
to a genus and inhering as a likeness is reducible to definition . 82 Thus, the way
in which the researcher comes to know what kind of thing the subject is and
what distinguishes it from others is through a process of division. In his De
animalibus , Albert shows how the method of division is to be applied to zoo¬
logical studies. Throughout his discussion, he insists that the zoologist must
avoid false and accidental divisions, for these will yield results that are not
convertible with the species and, therefore, fail to define it . 83 In particular, he
makes it clear that the method of dichotomous division used by the Platonists
will not do and he follows Aristotle in rejecting this method and setting out
rules for properly dividing animal species in a way that will result in valid
definitions . 84

Dichotomous division attempts to characterize a species by means of divid¬
ing a genus into two or, more properly, by one differentia at a time. This is
done at each successive stage of division until the species is identical with the
form to be defined. Definition, then, is constituted by reading back through
the divisions to find the successively greater genera in which the species par¬
ticipates . 85 Duck, on this method, might be defined by dividing the genus

81) “Et dicitur de omni problemate in genere, quoniam omne problems, vel est problems de
inesse, sicut quod est inesse ut sccidens, vel inesse ut genus, vel inesse ut proprium, vel inesse ut
diffinitio: et qusndo idonei sumus sd srtem syllogizsre sd ills qustuor, sumus potentes syllogi-
zsre omne problems, quod ex probsbilibus potest ostendi inesse.” Topica I, c. 2 (Ed. Borgnet
2:236s).

82) “Quia inesse ut differentia reducitur ad genus, et inesse ut idem reducitur ad diffinitionem.”
Topica I, c. 2 (Ed. Borgnet 2:236a).

83) “Constat enim, quod quarumdam rerum quae sunt species specificae, ultima differentia est
una tantum, et haec est convertibilis, sicut probatur in septimo primae philosophise: et quae-
cumque aliae assignantur differentiae, sunt superfluitates in plus existentes quam ipsa species
constituta per differentias: et tales superfluitates sunt, quibus non indigetur ad specierum consti-
tutionem.” De animalibus XI, tr. 2, c. 1 (ed. Stadler 13:780.28-34). See also Metaphysica VII,
tr. 4, c. 3 (ed. Colon. 16/2:370-72) and Quaestiones de animalibus K I, q. 7 (ed. Colon. 12:221).

84) David Balme provides a detailed study of Aristotle’s treatment of zoological division in “Aristo¬
tle’s Use of Division and Differentiae,” in Philosophical Issues in Aristotle's Biology, ed. Allan Gotthelf
and James G. Lennox (Cambridge, 1987), 69-89. Albert’s treatment in De animalibus XI, tr. 2,
c. 1-2 (ed. Stadler 13:780-92) generally follows, with a few additional examples and digres¬
sions, Aristotle’s Departibus animalium I (642b5-644al2) as analyzed by Balme.

85) Following Aristotle, Albert refers to dichotomous division as “dividing the genus through two
differentiae” {dividere genus per duos differentias). Here the differentiae are the two infimae species
into which the genus is divided by a single differentia, as when the genus aquatic bird is divided
into web-footed and non-web-footed. I generally follow Balme in characterizing dichotomy as
single-differentia division.

M. W Tkacz / Vivarium 45 (2007) 30-68

animal into aquatic and terrestrial, aquatic into footed and footless, footed
into biped and quadruped, and biped into web-footed and toe-footed. Then,
reading back through the divisions, duck is seen to be a web-footed, bipedic,
aquatic animal.

Such a method, Albert argues, is either invalid or incomplete, for it cannot
avoid accidental or false division unless it fails to include sufficient characteristics
to constitute a useful definition . 86 One reason is that dichotomy divides natu¬
ral kinds and, therefore, the same sub-kind may end up on both sides of a
supposedly exclusively divided genus. If animal is divided into aquatic and
terrestrial, then we cannot proceed to divided either side into footed and foot¬
less, for then some of each would appear on both sides. The division would fail
to show that being footed is an aquatic trait and, given that footed would not
thereby exclusively imply inclusion in a particular genus, the division would
not show what it is to have that trait in terms of what genus the subject
belongs . 87 Amother problem is that dichotomy cannot use differentiae which
indicate privations. Negative differentiae such as footless or featherless cannot
be divided further to provide useful definitions. One might try to divide foot¬
less animal into fish and snakes, but what is actually being divided is animal
and not footless: the genus really being “animal lacking feet.” So, footless as
such fails to allow for further division. While there are exclusive ways of being
footed, there are no infimae species of footless and, therefore, dichotomy can¬
not use negative differentiae . 88

Zoological division cannot be dichotomous, but must be by multiple dif¬
ferentiae at once, if it is to define the species under study in such a way that
the facts about the species can be established through quia demonstrations.
Following Ajristotle, Albert identifies three rules of valid division aimed at
avoiding the accidental divisions of the dichotomists. First, the differentiae by
which the genus is divided must be part of the definition or essential nature of
the species and not accidental. Even proper accidents will not be useful. A
certain species of octopus, for example, is characterized by having only one
row of suckers on each tentacle and can be identified as such. Yet, this is not
the proper definition of this species, for it is a proper accident attendant on
the narrowness of the tentacle that provides the true defining differentia . 89

86) DeanimalibusYl, tr. 2, c. 1 (ed. Stadler 15:780.19-781.7).

87) De animalibusYl, tr. 2, c. 1 (ed. Stadler 15:782.17-21).

88) DeanimalibusYl, tr. 2, c. 1 (ed. Stadler 15:782.22-786.10).

89) De animalibus XI, tr. 2. c. 1 (ed. Stadler 15:786.11-18). Albert does not here provide a zoo¬
logical example, simply repeating Aristotle’s mathematical illustration. The example here is given

M. W Tkacz / Vivarium 45 (2007) 30-68

Second, division must always be by proper opposites that will ensure that the
definiendum will fall exclusively on one side of the division. Here the zoolo¬
gist must be careful to avoid changing the jundamentum divisionis as when
dividing by differentiae that belong to non-opposed genera such as locomo¬
tion and coloration. For example, the dichotomous division of animal into
swimming and unpigmented is invalid. The proper division is of fish into
pigmented and unpigmented . 90 Thirdly, care must be taken to avoid acciden¬
tal division when dividing by operations common to the soul and body—that
is, by activities of the organism taken as a whole. One can validly divide ani¬
mal into sighted and non-sighted, but not into wild and domesticated. In
general, where there is a univocal name for a particular form, one cannot val¬
idly divide by a common operational differentia that would yield wild and
tame as species, as the name would show up on both sides . 91

To these three rules aimed at avoiding the sort of accidental divisions of the
dichotomists is added the fourth rule that provides an alternative method.
Albert explains that the reason why dichotomy is invalid is that its differentiae
are not continuous {immediatae ). The dichotomous division of animal into
winged and non-winged, winged into multi-colored and solid-colored, and
multi-colored into domesticated and wild fails in this way. Being multi-colored
is not a way of being winged, but of being colored (pigmented). Being domes¬
ticated is an accidental determination of being multi-colored, for nothing
about being determined as domesticated prevents the determination of the
same infima species by the opposite genus wild. Such difficulties are avoided
by the correct method of polychotomous division. The zoologist must apply
all the relevant differentiae to the genus together. By using several coordinated
series of dichotomous divisions, each of which divides with continuous dif¬
ferentiae, the zoological investigator will be able to secure valid and useful
definitions. Albert states the requirement like this:

It has been established, therefore, that many immediately opposing differentiae, which may
be attributed to the same division, ought not to be differentiae of one and the same thing,
for they are immediately opposing. Instead, there ought to be only one final differentia.

by Balme, Aristotle's De Partibus Animalium I and De Generatione Animalium I (Oxford, 1985),
114, based on Aristotle’s discussion at De partibus animalium IV 685b 15, discussed by Albert at
De animalibus XIV, tr. 1, c. 4 (ed. Stadler 15:1010).

90) De animalibus XI, tr. 2., c. 1 (ed. Stadler 15:786.19-32). The example is from Aristotle
(643a31-34); Albert gives no example here.

91) De animalibus XI. tr. 2., c. 1 (ed. Stadler 15:786.32-787.24).

M. W. Tkacz / Vivarium 45 (2007) 30-68

However, many differentiae attributed to multiple divisions can be of one and the same
thing, as winged, biped and being split-footed, because one division is into winged and
non-winged, another is of winged into biped and non-biped, and a third is of biped into
having a split-foot and not having a split-foot. 92

The zoologist does not begin with a very general simple concept such as ani¬
mal, but with the observed major genera such as bird or fish. These can be
immediately identified by various generic properties: bird, for example, is rec¬
ognized as winged, biped, beaked, and so on. These differentiae can then be
divided according to the various ways they appear in the genus. Following
Aristotle, Albert argues that these first-level generic differentiae are divided,
not by new differentiae but by the degree in which they appear in the various
sub-kinds. Thus, a genus defined by such differentiae is divided into various
species according to “the more and the less .” 93 All birds are beaked, but not in
the same way. Some have short, strongly curving beaks such as the hawk and
some have large, flat beaks such as the shoveler duck and some have long, thin,
pointed beaks such as the curlew. By dividing each of the generic characteris¬
tics in this way, one can come to know the differing sub-kinds of a certain
kind and to know what makes them different.

The zoologist need not restrict such divisions by degree to differentiae of the
major genera, but might also make certain dichotomous divisions which can
then be further divided by the more and the less. This will be valid as long as
the dichotomous divisions are continuous. Albert gives the example of the sub¬
genus water fowl which can be divided into toe-footed and web-footed. Each of
these species might be further divided by degree of toe-footedness (by length or
jointedness or shape of toe) and by degree of web-footedness (by shape or con¬
tinuousness of webbing). All these divisions would presuppose a division of bird
into aquatic and terrestrial and care in constructing the subsequent division of
each of these to preserve continuity . 94 Thus, for example, aquatic bird would
constitute a genus divisible into toe-footed and web-footed on the presump-

92) “Declaratum est igitur, quod multae differentiae immediate oppositae non debent esse dif¬
ferentiae unius et eiusdem rei, quae uni et eidem divisioni sint attributae, quia illae sunt immedi¬
ate oppositae, sed ultima differentia debet esse tantum una. Plures autem quae pluribus
divisionibus attribuuntur, unius et eiusdem rei esse possunt, sicut alatum et bipes et fissi pedis
existens, quoniam una divisio est in alatum et non alatum, et altera est alati in bipes et non bipes,
et tertia bipedis in fissum pedem habens et non fissum pedem habens.” De animalibus XI, tr. 2,
c. 1 (ed. Stadler 15:789.4-12).

93) De animalibus XI, tr. 2, c. 2 (ed. Stadler 15:789.150*

94) De animalibus XI, tr. 2, c. 1 (ed. Stadler 15:788.8-16) where Albert discusses simple and
complex differentiae.

M. W. Tkacz / Vivarium 45 (2007) 30-68

tion that by aquatic bird’ is intended bird whose primary “ground” (i.e., non¬
flying) locomotion is through wet-lands or on open water.

Using division to define the various forms of animals, according to Albert,
is not simply a matter of making a series of dichotomous divisions and then
“gathering” them together, in the Platonic manner, into a hierarchy of formal
features. Rather, definition by division proceeds by dividing off a kind by its
specific features and then differentiating them according to the degree they
exist in the morphological variants. Only in this way will the zoological inves¬
tigator be able to arrive at a final differentia that is convertible with the species.
The division then stands as an intelligible account of what was encountered in
observation,

because the proper and specific differentiae are known from the specific forms of things
and, therefore, the comparison will be the same between the forms of those things and the
form essentially and between the differentiae and the species rationally. Thus, just as the
specific form does not exceed the formed thing (that is, the formed thing under consider¬
ation) because the form is not found outside the formed thing nor the formed thing with¬
out form—for a cow is not found apart from the form of cow nor is the form of cow found
apart from a cow, and likewise with the ass and other animals—so it will be concerning a
specific differentia with respect to the species: it does not exceed species in the thing under
consideration nor does species exceed it. 95

If accidental division is avoided, then the final differentia will have precisely
the same extension as the species which is being defined. Thus, all the differ¬
entiae other than the final one are superfluous in the sense that each is a deter¬
mination of its predecessor and will imply it . 96 The zoologist, rationally
sorting-out the results of observation by means of non-accidental divisions
that are continuous and allow for the convertability of the final differentia
with the species, is articulating how the animal subject under study exists . 97
This is no less true for differentiae that are a difference of degree rather than
type of morphology:

95) “... quod differentiae propriae et specificae accipi habent ex formis rerum specificis, et ideo
eadem erit comparatio ipsius formae ad formam essentialiter et differentiae ad speciem rationaliter.
Et ideo sicut forma specifica non excedit formatum, scilicet formatum in suppositis, quia non
invenitur extra formatum nec formatum sine forma—non enim invenitur bos praeter formam
bovis nec forma bovis praeter bovem, et similiter de asino et huiusmodi—et sic erit de differen¬
tia specifica respectu speciei, quod ipsa non excedit speciem in suppositis nec species ipsam.”
Quaestiones de animalibusYl, q. 7 (ed. Colon. 12:221.21-32).

%) De animalibusXl, tr. 2, c. 1 (ed. Stadler 13:780.30).

97) MetaphysicaWW , tr. 4, c. 3 (ed. Colon. 16/2:371.1-11).

M. W. Tkacz / Vivarium 45 (2007) 30-68

For example, between one bird and another, which are of one and the same genus, there is
a difference according to the more and the less in the same nature, as, for instance, each is
winged, but one is more winged in that it has longer wings and the other is less winged in
that it has shorter wings.. . 98

The researcher, applying the results of measurement to the description of the
species, uses valid divisions yielding a final morphological differentia express¬
ing in a quantitative manner the morphology characterizing and distinguish¬
ing the species."

Clearly, Albert understood that Aristotle’s rejection of the Platonic method
of division by dichotomy was not a rejection of division itself as a dialectical
tool for scientific research. He not only carefully discusses all of Aristotle’s
rules for non-accidental division, but he also treats division in terms of the
goals of an Aristotelian research program. Albert, like Aristotle, does not see
division as a means by which a zoological systematics in the modern sense can
be constructed. He has no interest in providing, or even attempting to pro¬
vide, a complete taxonomy with all major and intermediate forms specified
and related. Both in his Aristotelian commentaries as well as in his own zoo¬
logical researches, division functions as a method of securing non-accidental
definitions of animal forms for the purpose of causal explanation . 100

The function of division, definition, is achieved by grouping differentiae
under genera as a means of providing a descriptive account of how various
animal features are associated in reality. The ultimate purpose of this is causal
explanation of the associations known to exist in animal species. Such expla¬
nation will be the dialectical manifestation of the causal connections among

98) “Verbi gratia, inter avem et avem quae sunt unius et eiusdem generis, est differentia quae est
secundum magis et minus in eadem natura, quoniam utraque est alata, sed una est magis alata
quae est longioris alae, et alia est minus alata quae est alae brevioris ...” De animalibus XI, tr. 2,
c. 2 (ed. Stadler 15:789.28-32).

99) Albert understood Aristotle’s notion of substantial form to be compatible with the notion that
species of a genus can differ only in terms of the relative magnitude of their morphology. Moreover,
Albert used this notion himself in his own zoological researches; see, for example, the various quan¬
titative specific distinctions in his De anirrutlibusXXlll, De falconibus (ed. Stadler 16:1453-1493).
On Aristotle’s views, see Lennox, Aristotle’s Philosophy of Biology, 160-181.

i° 0) Q n t he difference between Aristotle’s zoological method of division and modern biological
taxonomy see Pierre Pellegrin, Aristotle’s Classification of Animals: Biology and the Conceptual
Unity of the Aristotelian Corpus , tr. Anthony Preus (Berkeley, 1986) and David Balme, “Aristotle’s
Use of Differentiae in Zoology,” in Articles on Aristotle, vol. 1: Science, ed. Jonathan Barnes,
Malcolm Scholfield, and Richard Sorabji (London, 1975), 183-93.

M. W. Tkacz / Vivarium 45 (2007) 30-68

the associated features . 101 Albert realized, however, that the zoologist cannot
proceed directly from a series of divisions to causal explanations and, yet, divi¬
sion does give rise to demonstrations . 102 Assuming that the zoologist has at
hand a set of divisions of a subject-genus he wishes to study, Albert explains
how this yields demonstrations:

If we put this into the form of a demonstration, let us say that A, the middle term in such
a demonstration, is animal, because animality is the cause of the things inhering in an
animal; let B, which is the major term, stand for the attributes inhering in an animal; let C,

D, and E be the minor terms representing the species of animals, none of which fall under
the other genera but all of which are equally animal in their diverse ways. The proposed
attribute of whatever is thought to be in the particular nature ought to be demonstrated, as
if we said that C, D, and E are man, lion, and horse. It is clear from such a disposition of
terms, then, that the cause on account of which the attribute B is in this D, say a lion, is
because D is A. Similarly, it is clear why this same attribute B is in the others, namely C and

E, say man and horse: whichever species have it, do so on account of A. Thus, to syllogize
this, we would say: All A is B, all C is A, therefore all C is B. 103

The zoologist best makes use of the information provided through division by,
as Albert put it, expressing it in the form of a demonstration — that is, by
recasting the genera and species in such a way that the connections among
them become clear. Given a certain observed form under study, the researcher
begins with the most common genus relative to it. He then selects the mor¬
phological features which belong to the genus and those features which belong
to these. The major genus bird is one known from commonly accepted group¬
ings of animals. The researcher, collecting all the features common to every
bird, is then in a position to provide the reason why a given feature belongs to

10,) Posteriora Analytica II, tr. 4, c. 6 (ed. Borgnet 2:218a-b) and De animalibusYA, tr. 2, c. 2
(ed. Stadler 15:789-792).

i° 2) Priora Analytica I, tr. 6, c. 8 (ed. Borgnet 1:647-50).

,03) “Hoc autem si ad demonstrationem in terminis communibus debeat reduci, ponamus quod
A medium in tali demonstratione sit animal, quia animalitas est causa inhaerentium animali; B
autem (quod est major extremitas) sit ea quae inhaerent animali ut passiones; C, D, E minor
extremitas sint quaedam animalia in specialissimis considerata, quorum nullum est sub alio, sed
omnia coaequaeva sub animali accepta et diversa: de quolibet in propria natura particulari
accepto debet demonstrari passio proposita, sicut si dicatur, quod C, D, E, sunt homo, leo, et
equus. Manifestum igitur tali dispositione terminorum facta propter quid ut causam est B passio
in D hoc est, in leone: quia propter hoc quod D est A. Similiter autem manifestum est propter
quid idem B est in aliis, scilicet C et E quae sunt homo et equus: cuilibet enim illorum inest
propter A: sic enim syllogizatur: omne A est B, omne C [est] A, ergo omne C [est] B.” Posteriora
Analytica II, tr. 4, c. 6 (ed. Borgnet 2:218b-219a).

M. W. Tkacz / Vivarium 45 (2007) 30-68

the kinds which fall under the genus. Articulating this in the form of a quia
demonstration shows what the reason is in terms of the kinds and sub-kinds
revealed through the divisions. Essentially, then, the zoologist has rigorously
shown how the species under study exists in terms of the associations made
through the divisions.

The role of this sort of demonstration in research is illustrated by Alberts
own divisions of the various species of eagle in his De avibus . 104 Having estab¬
lished the common features of the major genus bird, he proceeds to consider
the various species according to common nomenclature. In each case, he pro¬
vides the various differentiae of the species that mark it off from other species
of birds as well as those that place the determined species in the same genus
along with other species. The eagle, for example, belongs in the same genus as
the falcon, for both in terms of morphology and behavior they are equally
birds of prey. Yet, they can be distinguished by, among other features, the rela¬
tive shape and length of wings: eagles being characterized by large broad wings
as opposed to the falcons narrow long wings. Albert goes on to group together
under the genus eagle various other differentiated features: massive oblong
beaks, large yellow feet, large broad wings, short tail feathers, etc. This pro¬
vides the basis upon which he can show why it is that the European sea eagle
has short tail feathers: it is because this bird is a sub-species of eagle and in the
species eagle short tail feathers are associated with other known features of
eagle morphology, also possessed by the sub-species, in a way they are not in
every species of the genus birds of prey. Albert also makes use of his divisions
to show, by similar quia demonstrations, that other species, such as the golden
eagle, are also of the same genus, for they have the short tail feathers charac¬
teristic of eagles. Likewise he is able to show why eagles of any species can
rightly be considered birds of prey, for they possess the features common to all
birds of prey: strongly curved pointed beak, taloned feet, etc.

Making use of divisions “in the form of a demonstration” in this manner
advances zoological knowledge in two ways. First, the zoologist who is able
to assign a feature to a species on the basis of the inclusion of the species in
the genus knows which kind possesses the feature per se. Second, once the
zoologist makes such an assignment on the basis of his divisions, he possesses
a greater knowledge of how the kinds are related to each other and how
they form a unity of kind . 105 Nonetheless, Albert is clear that this kind of

104) De animalibusXXIUy De aquila (ed. Stadler 16:1433-37).

105) Posteriora Analytica II, tr. 4, c. 6 (ed. Borgnet 2:219a-b); De animalibus XI, tr. 2, c. 2
(ed. Stadler 13:790.33-791.12).

M. W Tkacz / Vivarium 45 (2007) 30-68

intermediate explanation is not the same as causal explanation. It is a type
of quia demonstration that, along with the divisions upon which it is based,
provides the zoologist with a rigorous means of organizing the information
arising out of observation and measurement. It is to be carefully distinguished
from the propter quid demonstrations which, making use of this pre-
explanatory organization of zoological information, provide the causal
explanations . 106

C. Zoological Explanation

As already discussed, Albert admits that more than one type of explana¬
tory factor must be considered in seeking scientific knowledge of animal
forms. Yet, following Aristotle, he insists that the most fundamental type of
explanation, to which all other types are subordinate, is final cause. The
method of causal explanation in zoology, therefore, must be in accord with
this principle:

In the following books [on the parts and generation of animals], we will assign many causes
on account of the fact that, according to nature, there are many causes of generation and of
generated things, as shown in the second book of our Physica. We will state, for example,
the final cause, which is the cause of causes, and we will also state the efficient or moving
cause, which is the source of motion. This is because these two causes coincide with form
in everything developing toward form. Then, seeking to produce complete science, we
should distinguish which of these causes are primary causes — the causes of other causes, as
is the final cause—and which are secondary causes, receiving their causality from these
causes such that without these they are not causes. 107

The zoologist will be concerned to identify and demonstrate all the causes of
animal forms, including the material and the mechanical or efficient causes.
True and complete causal explanation, however, will be attained through

,06) Posteriora Analytica II, tr. 4, c. 7 (ed. Borgnet 2:222a).

,07) “Nos autem in sequentibus libris [de partibus et generatione animalium] multas causas
assignabimus, secundum quod multae sunt causae generationis et generatorum secundum
naturam, ut in secundo nostrorum ostensum est Physicorum : sicut verbi gratia dicemus causam
finalem, quae causa est causarum, et dicemus causam efficientem sive moventem, quae est prin-
cipium motus, quae in idem concidunt cum forma in omnibus mobilibus ad formam: et deinde
volentes facere s[c]ientiam perfectam distinguemus quae istarum sunt causae primae, quae sunt
causae aliarum causarum sicut est finis, et quae sunt causae secundae, quae ab illis habent cau-
salitatem suam, ita quod non causant sine ipsis.” De animalibus XI, tr. 1, c. 2 (ed. Stadler
15:765.28-38).

M. W. Tkacz / Vivarium 45 (2007) 30-68

demonstrations of the final causes of the forms known to exist through the
zoologist’s divisions . 108 Albert makes it clear that, because the material disposi¬
tions and the efficient operations present in animals are so for the sake of the
final cause of the animal, zoological explanation cannot be reduced to demon¬
stration of material and efficient causes alone. In this, he points out, the zoolo¬
gist is like the physician or other skilled practitioner:

So the physician, who is a natural scientist of a sort, does when he defines medicine through
sensible material, saying that it is the science of the healthy and the ill and those who are
neither, because it is the science by which the sensible dispositions of the human body are
known according to those things which heal and those which infirm. Likewise, the carpen¬
ter, who is the imitator of the natural scientist, when he defines knowledge of carpentry
through the form of his skill which involves sensible material. Both of them, in their defini¬
tions, give the final causes, on account of which the efficient causes produce in the sensible
material subject all they produce. The final and best intention is the cause on account of
which or final cause in the operations of nature and of the arts. The necessity that is in all
natural things is discovered in them in the same way as it is in the arts and medicine. 109

This analogy of the work of the natural scientist and that of the skilled practi¬
tioner indicates how Albert conceives of zoological explanation. Just as a phy¬
sician must know the material and mechanical dispositions of the body in
order to heal, so the zoologist will be concerned with the material components
and their efficient operations in so far as they are causes of the animal organ¬
ism as an ens completum. Zoological explanation will be primarily in terms of
the subject conceived as a terminus or perfection for the sake of which any
material is present and mechanical causes are operating. Material and efficient
causes, then, are secondary explanatory factors that cannot enter into the

i°8) “Videtur autem in omnibus mobilibus ad formam, sicut diximus quod causa prima quae
diffinitur esse ilia propter quam sint omnes aliae sit eadem cum forma, a qua sumitur rei diffini-
tio vera in omnibus rebus illis, quarum generatio est sustentatio in esse vero per naturam, quae
vere est forma rei.” De animalibus XI, tr. 1, c. 2 (ed. Stadler 15:766.1-5).

,09) “Sic enim facit medicus, qui particularis est physicus, quando diffinit medicinam per mate-
riam sensibilem dicens quod est scientia sanorum et aegrorum et neutrorum, eo quod ipsa est
scientia qua humani corporis dispositiones sensibiles cognoscuntur secundum ea quibus sanatur,
et secundum ea quibus infirmatur. Similiter autem facit carpentarius, qui imitator est physici,
quando diffinit per formam artis, quae concipit materiam sensibilem, scientiam carpentandi.
Ambo enim isti in diffinitionibus suis dicunt causas finales, propter quas faciunt causae efficien-
tes [faciunt] super materiam subiectam sensibilem omne quod faciunt. Ultimum enim intentum
et melius est causa propter quam sive finalis in operationibus naturae et artificii. Necessarium
autem quod est in rebus omnibus naturae, secundum eumdem modum invenitur in eis, secundum
quern invenitur in artibus et medicina.” De animalibus XI, tr. 1, c. 2 (ed. Stadler 15:766.11-28).

M. W. Tkacz / Vivarium 45 (2007) 30-68

explanation of the subject except relative to final cause or, as both Aristotle
and Albert often put it, with respect to that for the sake of which they enter
into the explanation at all . 110

Because scientific explanation is always a matter of demonstrating a neces¬
sity in the subject, Albert realizes that an account of zoological explanation
must indicate both the type of necessity that applies to zoological subjects as
well as that with respect to which the necessity applies. As already discussed,
Albert follows Aristotle in ascribing suppositional necessity to zoological sub¬
jects, explicitly contrasting this to the necessity of eternal things (in rebus
aeternis ), such as mathematical subjects. Unlike mathematical demonstra¬
tions, then, the demonstrations of the zoologist will always be on the supposi¬
tion (propter suppositionem) of the form to be explained- that is, they always
contain reference to the end and always proceed from effect to cause . 111

Final cause, however, functions in zoological subjects in more than one way
and, therefore, the necessity demonstrated in zoological explanations will be
of two types:

It is evident that all [zoological subjects] have two causes: namely, the “on account of
which” and the necessity of the material, which is necessary with respect to the supposed
end, as we have already said many times. Many things exist from such necessity as from a
cause: all generable things are produced by such causes, of which there are a great many. 112

Both of these explanatory factors involve necessity with respect to a final cause,
but in different ways. The material, with its attendant material characteristics,
is necessary on the supposition of an end (in suppositione finis). This, explains
Albert, is where the final cause is identical with the existence of the animal and
the material is necessary respectu esse. For example, certain material elements,
such as breathable air, are necessary to the animal in this sense. Yet the zoo¬
logist is not only interested in the existence of the animals, but also in their
specific morphologies (causas figurae membrorum ipsorum). Thus, final cause is
also the good of the animal for which certain morphologies and behaviors are

110) See, for example, Physica II, tr. 3, c. 7 (ed. Colon. 4/1:142.14-19).

110 De animalibus XL, tr. 1, c. 2 (ed. Stadler 15:761-768); see Wallace, “Albertus Magnus on
Suppositional Necessity in the Natural Sciences,” 120-25.

112) “Manifestum est igitur, quod omnes res tales habent duas causas, scilicet propter quid et
necessitatem materiae, quae necessaria est supposito fine, sicut iam saepius diximus. Multae
enim res sunt ex tali necessitate sicut ex causa: omnia enim generabilia ex talibus causis proce-
dunt, quae valde sunt multa” De animalibusXL, tr. 1, c. 3 (ed. Stadler 15:776.36-777.3).

M. W. Tkacz / Vivarium 45 (2007) 30-68

respectu bene esse . Camouflage markings, for example, are necessary for the
good of the animal in that they provide protection from predators . 113

Albert makes the same distinction in a somewhat different way in a digres¬
sion at the end of his treatment of zoological method . 114 He reminds the zool¬
ogist that the animal form discovered through division must not be considered
in a Platonic manner as separate from the material, but must be understood in
terms of the potencies of the animals material composition. He goes on to
explain that this is because the form is the end of a generative process ( finis
generationis) for which the material is suppositional^ necessary. The end in
this sense is identified with the being of the animal. There is also, however, the
end of the generated thing ( finis rei generatae) for which it is suppositionally
necessary that there be an animal of this kind. The end in this sense is identi¬
fied with the life of the animal or the function of its morphological parts.
Research is often aimed at explaining a particular form in terms of both types
of final cause. Albert makes a brief reference to the way morphological expla¬
nation in both senses might be part of a study of animal locomotion. An ani¬
mal need not have feet—after all, there are footless animals. Yet, if an animal
is to have feet, material of a certain kind is necessary. An animal need not have
feet of a certain shape—after all, animals are sometimes are born defective.
Yet, if the animal is to walk, then feet of a certain shape are necessary . 115

Explanation of the forms of animals, then, is a matter of demonstrating
what is necessary both ad esse and ad bene esse. Because the necessity in both
cases is ex suppositione finis , the end must be known prior to the process of
causal demonstration. This is the function of the descriptive phase of research
in which the animal forms are discovered through division and clearly articu¬
lated in quia demonstrations. The zoologist is then in a position to explain the
forms in terms of why they exist—that is, demonstrating that certain material
dispositions are necessary for their existence. The zoologist is also able to
explain the forms in terms of why they exist the way they do—that is, dem¬
onstrating that the forms are necessary for the animal to live as it does . 116

n3) De animalibus XI, tr. 1, c. 3 (ed. Stadler 15:777.16-28) where Albert contrasts these two
types of suppositional necessity with the absolute necessity (necessitas simpliciter ) of mathematics
and logic (in disciplinis ) and the necessity of force (necessitas faciens vim) considered in ethics; see
also Super I Sententiarum, dist. VI, A, art. 2 (ed. Borgnet 25:198b) and Metaphysical, tr. 1, c. 6
(ed. Colon. 16/1:220).

,,4) De animalibus XI, tr. 2, c. 4 (ed. Stadler 15:797.15-29).

U5) De animalibus XI, tr. 1, c. 3 (ed. Stadler 15:777.37-778.2).

1,6) DeanimalibusYl , tr. 1, c.3 (ed. Stadler 15:779.33-12) where Albert comments on Aristotle’s
sample zoological explanation of Departibus animalium I (642a31-b2).

M. W. Tkacz / Vivarium 45 (2007) 30-68

VI. Conclusion

The final five books of Albert’s monumental De animalibus contain the results
of his original zoological investigations arranged in the form of a traditional
medieval bestiary: a compendium of animal histories divided according to the
major genera (humans and quadrupeds, birds, aquatic animals, serpents, and
vermin) and listing species alphabetically . 117 While these are hardly complete
studies and are presented only in summary fashion, they nonetheless provide
evidence that Albert was engaged in the same type of zoological investigations
as the early Peripatetics. The details of his conception of an Aristotelian research
program, however, are found in his Aristotelian commentaries where he set out
at some length the nature and method of the natural sciences. James A.
Weisheipl has argued that the commentaries establish Albert’s great concern to
interpret Aristotle correctly as well as an obvious personal interest in explain¬
ing, correcting, and extending Aristotle’s research . 118 As argued here, more¬
over, the specific elements of Aristotle’s zoological research program are fully
set out and analyzed in these commentaries. Albert clearly understood the
need to mark off a self-contained subject-genus as a unified object of scientific
investigation and to show how animals constitute such a subject-genus. He
also conceived the study of this subject-genus as the pursuit of theoretical
knowlege in a specifically Aristotelian sense and offered a defense of the possi¬
bility of attaining such knowledge. Further, Albert identified natural finality
and suppositional necessity as principles of a science of animals. Finally, he care¬
fully set out in some detail a two-staged methodology for conducting animal
research making use of Aristotle’s methods of division and demonstration.

In an “Unsatisfactory Postscript” to his paper on the disappearance of Aris¬
totle’s biology , 119 James Lennox points out that familiarity with the libri de
animalibus , while possibly a necessary condition for pursuing the research
program found therein, was clearly not sufficient. The texts were certainly

117) De animalibus XXII-XXVI (ed. Stadler 16:1349-1598). For a discussion of the contents and
sources of these books see Kitchell and Resnick, 1-42.

118) Weisheipl argues that Albert defended Aristotelian naturalism from contemporary attempts
to platonize it in “Albertus Magnus and the Oxford Platonists,” Proceedings of the American
Catholic Philosophical Association 32 (1958), 124-39. Weisheipl also defends Alberts personal
commitment to Aristotelianism in “Albert’s Disclaimers in the Aristotelian Paraphrases,” Pro¬
ceedings of the International Conference on Patristic , Mediaeval and Renaissance Studies 5 (1980),
1-27. See also Fernand van Steenberghen, “Albert le Grand et l’Aristotelisme,” Revue Internatio¬
nale de Philosophie 34 (1980), 566-74.

119) Lennox, 123-24.

M. W. Tkacz / Vivarium 45 (2007) 30-68

known in antiquity and preserved and translated during the early medieval
period . 120 Yet, no one then took up the research task begun by Aristotle and
the first generation of his students. Lennox argues that what was needed in
addition to knowledge of Aristotle’s zoological treatises was an active and
intelligent researcher who shared the theoretical goals and methods outlined
in those treatises. Only such a researcher would be in a position to extend the
program of zoological investigation beyond the point reached by the early
Peripatetics. Lennox’s suggestion that Albert the Great was such a person is
fully justified by the contents of his Aristotelian commentaries, historically the
first scientific commentaries to be produced on the libri de animalibus. The
reasons for the fifteen-hundred-year dormancy in zoological research remain a
mystery and await further study. Nonetheless, the textual evidence presented
here supports the contention that the taste for and commitment to the
Aristotelian zoological research program was born again in Albert the Great.

l20 * David C. Lindberg, “The Transmission of Greek and Arabic Learning to the West,” in Science
in the Middle Ages, ed. David C. Lindberg (Chicago, 1978), 52-90.

BRILL

Vivarium 45 (2007) 69-92

VIVA

RIUM

www.brill.nl/viv

Averroes, Thomas Aquinas and Giles of Rome
on How This Man Understands

Brian Francis Conolly

Simons Rock College of Bard, Great Barrington, Massachusetts

Abstract

Giles of Rome, in his early treatise, De plurificatione possibilis intellectus , criticizes the
arguments of Thomas Aquinas against the Averroist doctrine of the uniqueness of the
possible intellect on the grounds that Aquinas does not fully appreciate the distinction
between material and intentional forms and the differences in how these forms are
generated. Nevertheless, like Aquinas, he argues that Averroes’ doctrine still results in
the apparently absurd consequence that homo non intelligit , i.e., the individual, par¬
ticular man, this man, does not understand. Giles, however, attempts to respond to
certain “radical” Averroists, who, in a bold and clever maneuver, affirm that homo non
intelligit. While Giles does effectively argue that homo non intelligit is not the opinion
of Averroes, he is unable to demonstrate the absurdity of homo non intelligit in a man¬
ner that would be convincing to the Averroists. This is because Giles, like Aquinas,
maintains that the intellect is a power of the soul, and thus has a different conception
of the relation between body and intellect than do the Averroists, who emphasize the
separateness of the intellect.

Keywords

intellect, Averroes, Averroism, Thomas Aquinas, Giles of Rome

Introduction

By the time Giles of Rome had written De plurificatione intellectus possibilis x in
about 1275, the Averroist doctrine of the separated and unique intellect shared
by all human beings had already been the source and focus of considerable

0 Giles of Rome, De plurificatione intellectus possibilis (Venice, 1550). There is a more recent
edition of this work, by Helga Bullotta Barrocco (Rome, 1957). This, however, hasnot been

DOI: 10.1163/156853407X183180

B. F. Conolly / Vivarium 45 (2007) 69-92

controversy for at least twenty years. It had already provoked official Condem¬
nation in 1270, and was the occasion for no less than two treatises written
expressly against it, by no lesser minds than Albertus Magnus 2 (1256) and
Thomas Aquinas 3 (1270), as well as being the subject of numerous ordinary
and quodlibetal quaestiones. Why, then, despite all this energy and attention
assembled against the doctrine, did Giles of Rome still find it necessary to
combat this Averroist doctrine at this relatively late date, with yet another
treatise expressly devoted to its refutation?

The answer can be found by examining a number of important texts writ¬
ten after 1270 that address the controversy. For, despite the Condemnation of
1270, and despite Aquinas’ powerful polemic in his De imitate , there is evi¬
dence of some doubt and dissatisfaction concerning the cogency of the argu¬
ments that had been leveled against the doctrine thereto. This dissatisfaction
can be found on the part of both the defenders and the critics of the Averroist
teachings.

Most important among the former is a set of quaestiones on Aristotle’s De
anima , which has come down to us anonymously, but is almost certainly the
work of Boethius of Dacia. 4 The bold Averroist author of these quaestiones ,
who defends the notion that the intellect must be separated from matter, dis¬
covered a way of undermining at a single stroke the most potent of Aquinas’
arguments against the doctrine. For Aquinas tried to show that a necessary
consequence of the Averroist doctrine of the unique and separated intellect is
the absurd proposition homo non intelligit , i.e., u this man , the individual par¬
ticular human being, does not understand”. The proposition is considered
absurd because it appears to be contrary to our most basic reflexively con¬
scious cognitive experience. Nevertheless, the anonymous author actually
affirms what Aquinas considers to be absurd: this man , indeed, does not under¬
stand ; rather, the separated intellect understands. This is not contrary to our

available to me. Fernand van Steenberghen, in his Maitre Siger de Brabant (Louvain, 1977), 118,
reports that in any case there are a number of problems with this Rome edition. Translations of
portions of Giles’ text in the present study are my own, and are based on the sixteenth century
Venice edition.

2) Albertus Magnus, De unitate intellectus contra Averroistas , ed. Bernhard Geyer in Opera Omnia
17:1 (Aschendorff, 1975).

3) Thomas Aquinas, De unitate intellectus contra Averroistas , ed. H. F. Dondaine, O. P. in Opera
Omnia 43 (Rome, 1976); On the Unity of the Intellect against the Averroists, trans. Beatrice Zedler
(Milwaukee, 1968).

4) Un commentaire Averro 'iste sur les livres I et II du Traite de Tame, ed. Maurice Giele, in Trois
commentaires sur le Traite de Tame dAristote , eds. M. Giele, F. van Steenberghen, and B. Bazan
(Louvain, 1971), 11-118.

B. F. Conolly / Vivarium 45 (2007) 69-92

reflexive experience, our anonymous author implies; rather, there is a tacit
demand in this affirmation that we carefully distinguish between our self-
awareness and self -intellection? Aquinas’ arguments have no dialectical force
against such an opponent, who, again, affirms homo non intelligit.

But even among the critics of the Averroist doctrine, that is to say, even
among those who otherwise are allied with Aquinas in the struggle against
Averroism, we again find texts expressing a certain dissatisfaction with Aqui¬
nas’ arguments. The most important among these texts is Giles of Romes
De plurificatione intellectus possibilis. For, although Giles, in this treatise, ulti¬
mately offers a number of original arguments that try to refute the Averroist
teaching of the unique and separated intellect, his attacks against the Aver-
roists are preceded by a lengthy critique—and attempted refutation as if on
behalf of the Averroists—of Aquinas. His critique of Aquinas is subtle, and
addresses a number of fundamental metaphysical distinctions that, Giles tac¬
itly claims, Aquinas has overlooked. The most important of these is the dis¬
tinction between intentional and material forms, and the differences in how
these forms are generated. Such distinctions allow Giles to give a much more
charitable reading of Averroes than Aquinas offers, and thereby enable Giles to
undermine the Angelic Doctor’s arguments. Giles’ critique of Aquinas is inter¬
esting, furthermore, in that it reveals something of the distance between Giles
of Rome and Thomas Aquinas on fundamental metaphysical issues. 5 6

But apart from his critique of Aquinas, Giles takes the battle against the
Averroists still further than Aquinas does. For, although Giles accepts and
adopts Aquinas’ strategy of showing that the Averroist doctrine leads to the
proposition homo non intelligit , he takes very seriously and tries to refute
the Averroists’ affirmation of that proposition. That is, unlike Aquinas, Giles
does not merely assume that homo non intelligit is an absurd proposition; rather,

5) Aquinas himself was aware of the distinction, and usually refers to our self-awareness in terms
of perception. Yet “to perceive” [percipere\ is a cognitive function that differs from “to under¬
stand” [intelligere], insofar as the latter is a comprehensive cognition, whereas the former is not.
On this distinction, especially as it plays out in Aquinas’ critique of Averroes, cf. Deborah L.
Black, “Consciousness and Self-Knowledge in Aquinas’s Critique of Averroes s Psychology”, Jour¬
nal of the History ofPhilosophy 31 (1993), 349-383.

6) This is interesting because Giles is too often regarded as a mere follower or defender of
Aquinas. There is, to be sure, extensive agreement between these two thinkers—and indeed, a
profound influence of the one upon the other. Nonetheless, there are important and significant
differences between the two at a fundamental level. This will be discussed, below, and is made
apparent by Giles in his opusculum. Cf. also, Edgar Hocedez’s Introduction to his edition of
Giles’ Theoremata de esse et essentia (Louvain, 1930).

B. F. Conolly / Vivarium 45 (2007) 69-92

Giles tries to prove it to be so. Giles thus tries to compensate for the dialectical
weakness of Aquinas’ approach, and thereby strengthens the case against the
Averroist theory of the intellect.

Furthermore, Giles’ discussion of and arguments against his Averroist
adversaries help to clarify that at the root of the controversy is a fundamental
tension between the universality demanded by and claimed for human reason
and the individuality and particularity of our intelligent reflexive experience.

In the course of Latin Averroism at Paris we find two distinct but correla¬
tive and competing approaches to the problem of reconciling this tension. On
the one hand, when the Latin Averroists adopt the notion that intellect must be
separated (and perhaps even unique), they are conceiving of the intellect in
such transcendent terms that allow them to assign a certain priority to the intel¬
lect’s claims for universality. The problem for the Averroists is then to explain
how such a transcendent intellect can nevertheless be construed as my intellect.

On the other hand are those opponents of the Averroists, such as Thomas
Aquinas and others, who understand reason or intellect not as unique and
transcendent, but as a power directly attributable to each individual, particu¬
lar human soul. They thus assign priority to the fact that this is my under¬
standing. The problem for them is to show how the intellect of this particular
human being can nevertheless achieve the universality that it lays claim to. 7 8

It is therefore possible to see the Averroists as concerned with offering an
alternative to the Aristotelian-Thomistic view of the relation between intellect
and human being. For, it appears, not all were convinced that the Thomistic
notion of the intellect as a power of the soul, which, as the form of the body,
is a kind of material form, could be sufficiently separate in order to perform its
intellective and comprehensive function. There is therefore some real philo¬
sophical pressure motivating the Averroist theory of intellect.

1. The Intelligible Species in Averroes’ Noetics

At the heart of the entire debate between the Latin Averroists and their critics
is the intelligible species. This is because most of the Latins who read Averroes’
Commentarium magnum in Aristotelis de anima libroP thought that the Com¬
mentator maintained that it is the intelligible species which provides the link,
as it were, between the unique, separated possible intellect and individual

7) See, for instance, Thomas Aquinas, Summa Theologiae I, Q. 76, art. 1.

8) Averroes, Commentarium magnum in Aristotelis de anima libros , ed. S. Crawford in Corpus
commentariorum Averrois in Aristotelem (CCAA), vol. 6, pt. 1 (Cambridge, Mass., 1953).

B. F. Conolly / Vivarium 45 (2007) 69-92

human souls in such a way as to account for how we can speak of understand¬
ing in connection with each particular individual human being. 9 The oppo¬
nents of the Latin Averroists argue that it cannot.

In their account of Averroistic noetics, 10 the Latins maintain that the agent
intellect causes the intelligible species to be made understood in act [ intellectum
in actu] . Prior to its activation by the agent intellect, the intelligible species is
merely an intention in the imagination and is something understood only in
potency. However, once the intelligible species has been thus activated, it then
informs or otherwise acts upon the possible intellect. The result of this union
between intelligible species and possible intellect is, on the one hand the specu¬
lative or habitual intellect [ intellectus speculative or intellectus qui est in habitu ]
or the speculatively understood thing [ intellectum speculativum ], depending
upon whether one is looking at this union from the perspective of the possible
intellect that has been made to be understanding in act, or from the perspective
of the intelligible species insofar as it informs the possible intellect.

However, because for Averroes, on this account at least, both agent and
possible intellects are unique and separated from the individual human soul,
Averroes can explain human understanding only in terms of the conjunction
of a speculatively understood thing [intellectum speculativum ] with the soul.
That is, it is only when there is some intelligible species, understood in act,
conjoined with the human soul that we can talk about understanding in con¬
nection with the individual human being. The intelligible species therefore
acts as intermediary between soul and separated intellect.

Such mediation is possible, this account of Averroes maintains, since
it is the same intelligible species that informs both possible intellect and
the imagination. Insofar as it informs the possible intellect, it is intellectum

9) The discussion must be qualified in terms of the difference between what Averroes actually
says and what the Latins thought he said. For, as will be discussed later, Averroes never intended
the intelligible species to be the link, as it were, that would allow communication of the power
of understanding from the unique and separated possible intellect to individual human beings.
Rather, for Averroes, that power is never completely removed from the human being insofar as
he or she is a human being. This is because, according to Averroes, the agent intellect is the form
for the human being as such. His discussion of the intelligible species and its “two subjects”, i.e.,
the possible intellect and the imaginative faculty, is introduced in order to solve a number of
problems concerning the speculative intellect. Again, however, it is not introduced in order to
explain how the power of understanding is communicated to us.

!0) What follows is a very compressed summary. For a much more detailed account of Averroes
pyschology, cf. Herbert A. Davidson, Alfarabi, Avicenna, and Averroes, on Intellect (Oxford,
1992). The essentials of Averroes’ psychology can be drawn from his Commentarium magnum in
Aristotelis de anima , III, Texts 4 and 5.

B. F. Conolly / Vivarium 45 (2007) 69-92

speculativum and understood in act\ insofar as it informs the imagination, it is
simply an intention in the imagination and understood in potency . Unlike the
possible intellect, however, the imagination is neither separated nor unique,
but is numbered together with the individual human souls.

1.1. Aquinas Arguments against Averroes

Aquinas offers many arguments intended to refute Averroes theory. However,
he developed basically two arguments that show that Averroes’ doctrine results
in the proposition homo non intelligit. These two distinct arguments appear in
substantially the same form in the various works in which Aquinas makes use
of them. 11 We shall use the versions that are found in the polemical De unitate
intellectuspossibilis contra Averroistas , §§65-66, since the presentation there is
especially clear and since that is the work Aquinas expressly devotes to com¬
bating the Averroist doctrine.

The first argument that Aquinas offers attempts to show that the intelligible
species that informs the phantasy and the possible intellect cannot provide the
union necessary for attributing the power of understanding to this man. This is

because the conjoining would not be according to a single principle, but according to
diverse principles. For it is clear that the intelligible species, in so far as it is in the phan¬
tasm, is potentially understood; but it is in the possible intellect in so far as it is actually
understood and abstracted from the phantasms. If, therefore, the intelligible species is not
the form of the possible intellect except in so far as it is abstracted from the phantasms, it
follows that the possible intellect is not in contact with the phantasms through the intelli¬
gible species, but rather is separated from them. 12

Since, then, there can be no joining between the possible intellect and the
phantasms, it is not possible that this man , Socrates, understands.

The second argument proceeds by saying that even if the possible intellect
and the phantasms could be conjoined by means of the intelligible species, this
would still not be able to explain the fact that this man understands.

n) Cf., for instance, De spiritualibus creaturis , art. II; Quaestiones disputatae De anima y art. II;
Summa theologiae, I, Q. 76, art. 1 (first argument only).

,2) Thomas Aquinas, De unitate intellectus , transl. Zedler, 50. Cf. the Latin text (ed. Dondaine,
p. 303b): “quia ista coniunctio non esset secundum aliquid unum, sed secundum
diuersa. Manifestum est enim quod species intelligibilis secundum quod est in fantasmatibus, est
intellecta in potentia; in intellectu autem possibili est secundum quod est intellecta in actu,
abstracta a fantasmatibus. Si ergo species intelligibilis non est forma intellectus possibilis nisi
secundum quod est abstracta a fantasmatibus, sequitur quod per speciem intelligibilem non
continuatur fantasmatibus, sed magis ab eis est separatus.”

B. F. Conolly / Vivarium 45 (2007) 69-92

For it is clear that through the intelligible species something is understood, but through the
intellective power, he understands something; just as also through the sensible species
something is sensed, but through the sensitive power he senses something. This is why a
wall in which there is color whose sensible species in act is in sight, is seen and does not see;
but an animal having the power of sight in which there is such a species, does see. Now the
aforesaid union of the possible intellect with man, in whom there are phantasms whose
species are in the possible intellect, is like the union of the wall in which there is color with
the sight in which is the species of its color. Therefore, just as the wall does not see, but its
color is seen, so it would follow that man would not understand but that his phantasms
would be understood by the possible intellect. 13

In other words, any conjoining of possible intellect and phantasms would
not be sufficient to communicate the power of understanding to this man. The
best that this kind of union could result in for this man , Socrates, would be to
make him an object of understanding for the intellect; it would not, however,
make him to be an understanding subject. Because the power of understanding
does not and cannot communicate to Socrates when the possible intellect is
posited to be a unique and separated substance, it follows, according to Aqui¬
nas, that under Averroes’ theory, this man does not understand.

1.2. Giles Critique of Aquinas

That Giles’ opusculum presents a significant development over Aquinas’ argu¬
ments against the Averroists has been altogether missed by recent scholars. In
fact, Giles’ treatise has been regarded as something of an enigma. For after a
very detailed and charitable account of Averroes’ theory, he then offers some
rather powerful criticism of Aquinas’ arguments against the Commentator.
Indeed, the first half of the treatise could be read as a compelling defense of
the Averroist theory. However, it is the next section of the treatise that has
caused confusion. For it is there that Giles begins his own refutation of Aver¬
roes. The problem, however, is that he appears to be simply applying the same

13) Thomas Aquinas, De unitate intellectus, transl. Zedler, 50-51. Cf. the Latin text (ed. Don-
daine, pp. 303b-304a): “Manifestum est enim quod per speciem intelligibilem aliquid intelligi-
tur, sed per potentiam intellectivam aliquid intelligit; sicut etiam per speciem sensibilem aliquid
sentitur, per potentiam autem sensitivam aliquid sentit. Unde paries in quo est color, cuius spe¬
cies sensibilis in actu est in visu, videtur, non videt; animal autem habens potentiam visivam, in
qua est talis species, videt. Tabs autem est predicta copulatio intellectus possibilis ad hominem,
in quo sunt fantasmata quorum species sunt in intellectu possibili, qualis est copulatio parietis in
quo est color ad visum in quo est species sui coloris. Sicut igitur paries non videt, sed videtur eius
color, ita sequeretur quod homo non intelligent, sed quod eius fantasmata intelligerentur ab
intellectu possibili.”

B. F. Conolly / Vivarium 45 (2007) 69-92

arguments of Aquinas, albeit in a somewhat different form. Fernand van
Steenberghen’s assessment of this section of the treatise is typical:

Mais la maniere dont il critique les arguments de S. Thomas dans le troisieme partie tout
en les adoptant dans la quatrieme, est deconcertante. II pretend corriger les demonstrations
de son maitre; mais, en fait, on a l’impression qu’il repete tout bonnement S. Thomas sans
rien ameliorer. 14

This view is shared by Dales, Egenter, and others. 15

Nevertheless, Giles himself believes that his arguments differ from those of
Aquinas, albeit subtly. As will become clear, Giles’ account of Averroes’ doc¬
trine (and consequently his arguments against the Commentator) differ from
those of Aquinas on two fundamental points. First, Giles carefully distin¬
guishes between the manner in which material forms and intentional forms
inform their respective subjects. This distinction is completely overlooked by
Aquinas in his account of Averroes’ doctrine. Secondly, unlike Aquinas, Giles
subscribes to the “indifference” theory of universal, that is, that the intelligi¬
ble species in itself is neither universal nor particular, but takes on universality
only insofar as it informs the possible intellect. Although the strategy is the
same (namely, to show that Averroes cannot account for the communication
of the power of understanding to the individual human being), the underlying
metaphysics differ; and therefore, the arguments differ.

1.2.1. Whether the Same Species Informs Intellect and Imagination
In the first place, then, Giles argues that, contrary to Aquinas, one and the
same species can inform both the intellect and the phantasm, given that the
possible intellect is separated. Furthermore, although he agrees with Aquinas

14) Cf. Van Steenberghen, Maitre Siger de Brabant , 118.

15) Cf. Richard C. Dales, The Problem of the Rational Soul in the Thirteenth Century (Leiden,
1995). Dales (p. 163) determines that Giles treatment of the matter adds nothing philosophi¬
cally to Aquinas’, but makes improvements only in presentation. This view is shared by Richard
Egenter, Die Erkenntnispsychologie des Aegidius Romanics (Regensberg, 1925), 31-32, who pro¬
vides a rather complete summary of the whole of Giles’ treatise (14-32). Neither Dales nor
Egenter takes seriously Giles’ own claim that all of the arguments directed against the Commen¬
tator hitherto are inadequate, or Giles’ view that his own stand against Averroes really does differ
from that of his predecessors, albeit subtly. Indeed, as we shall see, Giles’ arguments advance the
discussion considerably beyond where Aquinas left it in his De unitate intellectus. Cf. Van Steen¬
berghen, Maitre Siger de Brabant, 118, for further references. In any case, none of the five addi¬
tional arguments that Giles himself offers against Averroes bears any resemblance to any of
Aquinas’ arguments.

B. F. Conolly / Vivarium 45 (2007) 69-92

that this union of phantasy and intellect is not sufficient to account for the
power of understanding communicating with this man, he rejects Aquinas
argument as insufficient.

Giles’ reasoning for why the same species can inform both the intellect and
the imagination is based upon two points: (1) that intentional forms are not
numbered according to the enumeration of the subjects that they are in; and
(2) that a spiritual nature can be together with a corporeal nature in such a
way that the corporeal nature does not recede or otherwise yield itself.

The first point is the more important, as it aims at showing that the same
species can inform the intellect and the phantasy. The argument proceeds by
contrasting intentional forms with real forms. While it is obscure in parts, it
seems to go as follows: 16 Every material thing is composed of matter and form
and is subject to generation and corruption. Now, in the generation of the
thing, it is not the form that is generated, but rather the composite, of which
the form is a part (the other part being the matter). Thus, for instance, it is not
white or black that is generated, but rather the white stone or the black stone.
Now the end of generation is being; so whatever being the form has must be
attributed to the composite (otherwise it would be a self-subsistent and so not
a material form).

And because such being is attributed to the composite, it follows that the numerical diver¬
sity in forms would have to be from the diversity of subjects. Since, therefore, it is proper
for every real form that it is not generated (whereas the composite is), so in every real form
it so happens that it is numbered by the enumeration of subjects. 17

Giles further explains that the reason the whole composite is generated (and
not merely the form) is because the whole composite generates. All material
forms are educed from the potency of matter, and for whatever is educed from
the potency of matter, the whole is educed from the whole, and the part from
the part.

This last point becomes more clear when it is compared to the generation
of the intentional form. The intentional form is generated when the form
alone acts upon some cognitive faculty—sense, imagination, intellect. Matter,
and hence also the composite, play no role in the generation of the intentional

16) Cf. Deplurificatione , 56v A-E.

17) De plurificatione , 56v C: “Et quia huiusmodi esse composito attribuitur sequitur quod diver-
sitas numeralis in formis ex diversitate subiectorum esse habeat, cum igitur hoc sit proprium
omni forme reali quod ipsa non generatur, sed compositum, ideo in omnibus formis realibus
competit numerari numeratione subiectorum.”

B. F. Conolly / Vivarium 45 (2007) 69-92

form. What generates a composite, and so what brings the real form into
being, is another composite. Therefore, in a sense, the composite generates the
real form. But in intentional things, only the intention or the intentional form
is generated, because the form alone generates.

Giles does not consider that it may be objected that because the intentional
form is not self-subsisting, it, too, is in a subject, and consequently is part of a
composite that is generated, say, the composite of the intentional form and the
imaginative faculty. Thus it would be wrong to say that only the intention is
generated, as opposed to saying that the composite or the whole is generated;
and so it would be wrong to say that the intentional form, unlike the real
form, is not numbered by the enumeration of the subject that it is in.

This objection may be ignored by Giles on the grounds that there must be
in any case some identity between the real form and the intentional form. If
the intentional form were numbered by its subject, then it would be numeri¬
cally distinct from the real form of which it is the image; and so there would
be no identity of real and intentional form. There would then be, however, an
unbridgeable chasm between reality and the cognitive powers; in which case
our cognitive powers would have no grounding in reality and so the possibility
of knowledge of the material world would be completely undermined.

Giles may be alluding to this identity between intentional form and real
form when he compares the likeness of the generating thing to the thing
generated.

For we see in the generation of real things, that the generated thing does not perfectly fol¬
low the likeness of the generating thing, because, since such likeness is by the unity of form,
the generated thing does not perfectly attain the unity of form of the generating thing. For
the action is not attributed to the form, but to the composite, though the form is the cause
of its action. 18

In other words, with respect to real forms, the form in the generating thing
and the form in the thing generated are not one, even if they are similar. They
are not one because the action of generating is attributed to the composite,
and not to the form. When is it the form alone that is acting, then the form of
the generating thing and of the thing generated are one, as in the case of inten¬
tional things.

18) De plurificatione , 56v D-E: “Videmus enim in realium generatione, quod genitum non per-
fecte sequitur similitudinem gignentis, quia cum talis similitudo sit per unitatem forme non
perfecte assequitur genitum gignentis unitatem forme, ex eo quod ilia actio non attribuitur
forme sed composito licet illius actionis forma sit ratio.”

B. F. Conolly / Vivarium 45 (2007) 69-92

But in the generations of the intentional things, the generated thing perfectly follows the
likeness of the generating thing, just as the likeness in the mirror perfectly follows the like¬
ness of the generating thing. 19

The unity of form between the generating thing and the intentional form thus
provides the identity between real form and intentional form necessary for the
possibility of knowledge.

This can be illustrated by supposing that Theophrastus is perceiving Aristo¬
tle and his son Nichomachus. Now the essential form of Nichomachus (and
by this we mean the form and the matter, apart from the existence of Nich¬
omachus) is similar to the essential form of Aristotle, which generated the
essential form of Nichomachus; but they are not exact. Aristotle and Nich¬
omachus do not obtain a perfect unity of form. However, the real form of
Aristotle and the intentional form produced by it in the phantasy of
Theophrastus (and here we mean the “form of the part”, i.e., the formal part
of the essential form, apart from the matter) are exactly the same. They do
achieve a perfect unity of form. That is, there is but one form.

To summarize Giles’ position thus far, then, real forms are numbered by the
enumeration of their subjects, because real forms are parts of composite mate¬
rial substances and it is the substance that is generated, not the form. Further¬
more, since being is the end of generation, and since it is the composite that is
generated, whatever real being the form has it owes to the composite; and
from a diversity in being—the end of generation—numerical diversity is
assumed. Real forms differ in being as the composite of which they are parts
differ; thus real forms are diverse as the composites of which they are a part are
diverse. Therefore, real forms are numbered according to the enumeration of
their subjects.

On the other hand, because with respect to intentional things only the
form is generated—there is no composite—it follows that they are not num¬
bered according to the enumeration of the subjects that they are in.

Because real forms and intentional forms differ with respect to how they are
numbered, Giles can draw several conclusions. First, with respect to real forms:
(1) two real forms in the same subject must differ in species as well as number,
and (2) one and the same real form cannot be in two different subjects. 20

19) De plurificatione , 56v E: “In generationibus intentionalium, genitum perfecte consequitur
similitudinem gignentis, sicut similitudo in speculo genita perfecte consequitur gignentis simili-
tudinem.”

20) Giles (Deplurificatione , 56v A) also cites Aristotle in support of his conclusions. Cf. Aristotle,
Metaphysics V, 1016a 17-24.

B. F. Conolly / Vivarium 45 (2007) 69-92

Second, because intentional forms are not numbered according to their sub¬
ject the converses of (1) and (2) hold for intentional forms: (3) two intentions
differing only in number can be in the same subject, and (4) one and the same
intention can be in two different subjects.

The last point, that one and the same intention can be in two different
subjects, is of course the very thing that Giles sought to prove in this discus¬
sion. For, having shown this, he will be able to assert that the same species can
inform the intellect and the imaginative faculty. 21 However, before he can

21) Alain de Libera, La querelle des universaux (Paris, 1996), 225, points out that Giles was
identified throughout the fourteenth century with precisely this view, which maintains the iden¬
tity of the species in the imagination and the species in the intellect, in such a way as to render
useless the intervention of the agent intellect as an abstractive causal agent. The theory of univer¬
sal from which it stems is actually a version of the “indifference” theory of universal, which
originated with the Islamic philosophers, beginning with Alfarabi. (De Libera, be. cit.) On Giles’
views on universal and the agent intellect, see Edgar Hocedez’s Introduction to his edition of
Giles’ Theoremata de esse et essentia (Louvain, 1930), especially pp. 74-82. Egenter (op. cit.,
pp. 50-55, 61-67) provides a discussion of Giles’ views on universal that is useful especially for
the numerous references to several of Giles’ works wherein Giles himself characterizes the view
on universal that he adopts as being the “common” view. The attribution to Giles of the view
that the action of the agent intellect is non-abstractive is challenged by Carey J. Leonard, “A
Thirteenth Century Notion of the Agent Intellect: Giles of Rome”, New Scholasticism 37 (1963),
327-358. Leonard shows that when considering the relation between the understood thing and
the material thing, Giles does indeed make use of a theory of abstraction along the lines of Aris¬
totle and Averroes. Leonard does not, however, consider the unity of form among the really
existing nature, the intention in the imagination and the intelligible species in the intellect,
which is central to Giles’ discussion in Deplurificatione. This must remain problematic for Giles,
for, since the really existing nature and the intention in the imagination have a unity of form—a
unity of form made possible because the material conditions of the really existing nature are not
impressed upon the imagination nor do they in any way contribute to the generation of the
intentional form informing the imagination, the intention in the imagination appears to be no
less abstract than the thing understood by the intellect. On this issue, cf. also Valeria Sorge,
“L’astrazione nella Gnoseologia di Egidio Romano”, Rivista di Fibsofia Neoscolastica 72 (1980),
670-680. Furthermore, Leonard’s reference to Averroes might not be all that helpful. This is
because a similar problem arises for the Commentator himself, who is sometimes ambiguous
with respect to the function of the various internal senses. At times these have the role of merely
re-presenting the material individual to the intellect in order that it may be abstracted. At other
times, however, especially in his Epitome of the Parva naturalia , Averroes appears to be following
Galen and gives us a much more materialistic account of abstraction and intellection. Cf. Aver¬
roes, Epitome of Parva naturalia , transl. H. Blumberg (Cambridge, Mass., 1961), esp. 23-27. The
internal senses are seen performing successive abstractive functions themselves, removing the
rind, as it were, from the material thing to get to the thing to be understood at the core. In these
cases, it is unclear what abstractive role is left for the agent intellect to perform.

B. F. Conolly / Vivarium 45 (2007) 69-92

embark upon his critique of Aquinas’ arguments against Averroes, he needs to
make clear the second point mentioned above, namely that a spiritual nature
can be together with a corporeal nature in such a way that the corporeal nature
does not recede or does not yield itself. Now the import of this point seems to
be in showing that the unique separated intellect of Averroes’ doctrine can be
conjoined with the imaginative faculty of this man in such a way that the
intellect and the phantasy remain distinct. For, if they did not remain distinct,
then because the intellect (a spiritual nature) is unique, it would not be the
case that this man understands. He does not seem to offer any argument for
this point, except, perhaps, only to point out that a corporeal nature cannot
be together with another corporeal nature. Thus, presumably, because a spiri¬
tual nature is not a corporeal nature, it is possible that a spiritual nature can be
together with a corporeal nature. 22

1.2.2. Giles Refutation of Aquinas' First Argument

With these two points established, namely, that intentional forms are not
numbered by the numeration of the subjects that they are in, and that a spiri¬
tual nature can be together with a corporeal nature in such a way that the
corporeal nature does not yield itself or recede, Giles can conclude that one
and the same intelligible species can inform the intellect and the imaginative
faculty or phantasy. Giles strengthens this claim by explaining that what inten¬
tional forms (and so, intelligible species) are numbered by are those things the
cognition of which they lead to. The intelligible species informing both the
intellect and the phantasy is one and the same because it leads to the cognition
of the common nature.

And since the species in the intellect leads to the cognition of the nature without material
conditions, whereas the species in phantasy leads to the cognition of the nature with mate¬
rial conditions, <then> just as one and the same nature is what is considered, with material
conditions and without them, so also one and the same species can be what informs the
phantasy and the intellect. 23

Having established that one and the same species can at the same time inform
both intellect and imagination, it is a short step to dismiss the first of Aquinas’

22) Cf. Deplurificatione , 56vE-F.

23) De plurificationey 56v G-H: “Et quia species in intellectu ducit in cognitionem nature
sine conditionibus materie, species autem in phantasia ducit in cognitionem nature cum conditionibus
materie sicut una et eadem natura est que consideratur cum conditionibus materie et sine illis
ita et species una et eadem esse poterit que informat phantasiam et intellectum.”

B. F. Conolly / Vivarium 45 (2007) 69-92

arguments made against the Commentator, that is, the one showing that one
and the same form cannot inform both the phantasm and the intellect. For,
because the informing intelligible species is in either case the same form, the
intellect can be united with the imagination as Averroes claims. Aquinas' argument
sought to establish that the informing species were distinct because of the
distinction between being understood in act and being understood in potency.
Giles’ argument, in effect, reminds us that whether the species is understood
in act or in potency is determined by the subject: as it informs the phantasm,
it is understood in potency; as it informs the possible intellect, it is understood
in act. And because intentional forms are not numbered by the enumeration
of their subjects, there is no reason to think that the species informing the
phantasm and species informing the intellect is not one and the same.

1.2.3. Giles Refutation of Aquinas' Second Argument

Recall that in his second argument, Aquinas tries to show that even if the pos¬
sible intellect could somehow be conjoined with the imaginative faculty of this
man y this man would still not understand because, according to Averroes’ the¬
ory, the intellect and phantasm would be united only as understanding subject
with understood object, or as perceiving subject with perceived object. Just as
the power of seeing is not communicated to the wall when the color in the
wall is seen, so the power of understanding is not communicated to this man
when the intelligible species in this man is understood.

Giles believes that Aquinas errs in this argument when he thinks that the
union achieved between intellect and phantasy is just like the union achieved
between the eye and a wall. This assumption is false, because the intelligible
species informing the intellect is numerically the same as the intelligible spe¬
cies informing the phantasm in the phantasy. They are numerically the same
because intelligible forms are not numbered by the enumeration of the sub¬
jects that they are in, for the reasons adduced earlier. On the other hand,
material forms, such as color, are so enumerated. Thus, the color informing
the eye—making the eye to be really green, for example—is numerically dis¬
tinct from the color informing the wall. The reason that the power of sight is
not communicated to the wall is because of the numerical diversity of the
informing forms.

The falsity of the first argument, which argues that a man does not understand (because the
intellect is not united with us in a union through which we men can understand, because
there is no more union from man and the intellect than from the wall and the eye), is
apparent enough. For it argues from a false analogy. For, as was shown (when the position

B. F. Conolly / Vivarium 45 (2007) 69-92

of the Commentator was explained), one and the same species informs the phantasy and
the intellect. Therefore the form of the intellect was the form in us, because of which the
act of understanding was communicated to us. But one and the same species does not
inform the eye and the wall; therefore the form of the eye is not the form in the wall,
because of which the act of seeing is not communicated to the wall. 24

Aquinas thus applies a false analogy in his thinking. The relation between the
wall and the eye is significantly different from the relation between the intel¬
lect and the phantasm. There is unity and identity of the intelligible species
informing the intellect and the phantasy, whereas there is a plurality and diver¬
sity between the forms informing the wall and the eye, respectively. Aquinas,
therefore, has not provided sufficient reason for thinking that the power of
understanding cannot be communicated to this man. 15

1.3. Giles' Own Refutation of Averroes

Giles actually offers several arguments against Averroes, but in keeping with
the theme of this discussion, we shall consider only that one which directly
addresses the Commentator’s account of how this man understands. We have
already seen that Giles is satisfied that the intelligible species can be the cause
of some unity between the separated possible intellect and the phantasms.
However, he does not think that this union is sufficient for the power of
understanding to be communicated to this man. He is thus in agreement
with Aquinas to this extent, that Averroes cannot account for the power of
understanding convening upon this man. His argument, of course, must be
different.

24) De plurificatione, 58r H-58v A: “Nam prima ratio que arguit hominem non intelligere, ex eo
quod intellectus non unitur nobis unione per quam possumus homines intelligere, quia non plus
fit unum ex homine et intellectu quam ex pariete et oculo, satis apparet falsitas huius rationis,
quia arguit ex falsa imaginatione. Nam sicut ostensum fuit, quando declarata fuit position com-
mentatoris, una et eadem species informat phantasiam et intellectum. Et ideo forma intellectus
erat forma in nobis, propter quod actus intelligendi nobis communicabatur, sed una et eadem
species non informat oculum et parietem, ideo forma oculi non est forma in pariete propter
quod actus videndi non communicatur parieti.”

25) Giles (Deplurificatione , 58r G) considers and rejects as puerile a way of responding to Aqui¬
nas’ second argument that he attributes to unnamed Averroists. Like him, they claim that the
argument is based upon a false analogy. They think the problem is that the Thomistic argument
draws an improper analogy between the wall and this man. The problem, they claim, is that the
reason that the wall does not see when it is seen is because it does notpossess a cognitive power;
whereas the individual man does possess a cognitive power, because of which this man under¬
stands when he is understood.

B. F. Conolly / Vivarium 45 (2007) 69-92

He argues that in order for the power of understanding to be communi¬
cated to us, it is necessary that the intelligible species be in us, but, that it be
in us insofar as it is intelligible in act.

Therefore, if we understand because the species (which is said to be the form of the intellect
according to the position touched on above) is in our imagination, it is necessary that the
intelligible species be in us and insofar as it is intelligible by us. But this because it is in act,
and acts as such. 26

The problem, however, is that the intelligible species is in act only insofar as it
is in the possible intellect; insofar as it is in us, it is merely intelligible in
potency.

Since, therefore, the cause of understanding is the intelligible species with respect to this,
that we understand through the intelligible species, it is necessary that the intelligible spe¬
cies be in us. For the species that is in the phantasy (which is the form of the intellect) is in
us. Nevertheless, insofar as it is intelligible, it is not in us, but it is in the intellect. 27

On the surface this does indeed sound like Aquinas’ first argument. The
difference, however, pertains to that for which the argument is offered as a
reason. Aquinas makes use of the difference between being understood in act
and being understood in potency in his claim that the intelligible species in
the imagination and the intelligible species in the intellect are numerically
distinct. That is, for Aquinas, the fact that one is understood in potency and
the other is understood in act is indicative of the numerical distinctness of the
species involved. Giles, on the other hand, has argued that this difference does
not indicate a numerical distinction with respect to the informing species.
Rather, Giles has argued that the same species can inform both intellect and
imagination. What this argument claims instead, is that this union is not
sufficient to account for the communication of the power of understanding.

The reason for why the act does not communicate, according to Giles, is
as follows:

26) Deplurificatione , 59r A: “Si igitur ex eo quod species est in imaginatione nostra, que dicitur
esse forma intellectus secundum positionem pretactam ad hoc quod intelligeremus, oportet
quod species intelligibilis esset in nobis et secundum quod esset intelligibilis nobis, sed illud
quod est in actu, et ut sic agit.”

27) De plurificatione , 59r A-B: “Cum igitur ratio intelligendi sit ipsa species intelligibilis, ad hoc
quod per speciem intelligibilem intelligamus, oportet quod species intelligibilis sit in nobis quia
species que est in phantasia que est forma intellecus est in nobis tamen secundum quod est intel¬
ligibilis non est in nobis sed est intellects”

B. F. Conolly / Vivarium 45 (2007) 69-92

The reason for this is that what is in potency is not acting. It is therefore necessary that the
agent be in act. Furthermore, because the agent does not act indifferently, whatever the act,
but proportionately, not only is the agent is in act, but the agent acts only to the extent it
is in act. 28

Thus, even though it is the same form that informs both imagination and
intellect, it is only in the intellect that the form is understood in act; and thus
it would be only the separated intellect that understands in act. If the power
of understanding communicated with the individual man, by means of the
form informing the imagination, that form would likewise have to be under¬
stood in act; but it is not. Rather, it understood merely in potency. 29

Giles’ difference with Aquinas on the refutation of Averroes is subtle, as
Giles himself recognizes. 30 Both argue that the power of understanding cannot
be communicated to the individual man within the Averroistic framework.
Aquinas argues that this is because the intention informing the imagination
and the intention informing the intellect are numerically distinct, and this,
because the one is understood in potency and the other understood in act.
Giles, however, claims that one and same intention informs both intellect and
imagination; but even so, this is not sufficient for the power of understanding
to be communicated to the individual man.

In other words, Aquinas has argued that no such conjunction is possible;
but even if it were possible, it would not be sufficient to account for the com¬
munication of the power of understanding to the individual, particular man.
Giles, contrary to Aquinas, argues that such a conjunction is indeed possible,
but even though it is possible, it still cannot account for the communication of
the power of understanding.

For both Giles and Aquinas the reason why the power of understanding
cannot communicate to the individual particular man is because the species,
insofar as it informs the imagination is understood in potency, whereas insofar
as it informs the intellect, it is understood in act. Giles’ difference with Aquinas,

28) De plurificatione , 59r A: “Cuius ratio est quia quod est in potentia non agit. oportet igitur
quod agens est in actu. Iterum quia agens non agit indifferenter quemlibet actum, sed propor-
tionatum, non solum agens est in actu, sed agens in tantum agit in quantum est in actu.”

29) The argument appears to be echoing the indifference theory of universal. In response to
a later argument by Aquinas, Giles explains that even among the hierarchy of separated
intelligences, there is no univocal intelligible species among the distinct orders. Rather, the
higher intelligences understand the species more universally than is understood by the lower
species. Cf. De plurificatione , 60rE-G.

30) De plurificatione , 59r C.

B. F. Conolly / Vivarium 45 (2007) 69-92

however, lies in the notion that the individuality or universality in question is
entirely attributable to the respective subjects informed by the species. The
form is received proportionately, that is, according to how the subject can
receive it. The imagination can receive the form only as individual and under¬
stood in potency, the intellect, however, receives it as universal and understood
in act. It is nevertheless numerically one and the same form informing both,
contrary to what Aquinas maintains. Individuality and universality, which is
to say, being understood in potency and being understood in act, therefore,
for Giles, stand outside the form itself, as it were. For Aquinas, they do not,
and this is, consequently, sufficient for him to establish the numerical diversity
of the informing species.

2. Homo non intelligit

For Giles, no less than for Aquinas, to say that this man understands means to
attribute the power or act of understanding to this man. That is, for them, to
say that this man understands is to say that the intellective power is properly
attributed to Socrates, and not merely to some separated intellect. Both Giles
and Aquinas have argued that even if Averroes can account for some connec¬
tion between the separated intellect and Socrates, this connection is not
sufficient for the act or power or understanding, which properly belongs to the
intellect, to be communicated to Socrates in such a way that this power can be
properly attributed to him. Now, supposing that their arguments are sound,
the proper evaluation of their arguments as refutations of Averroes still requires
that it be determined whether the Commentator is indeed committed to
properly attributing the power of understanding to this man.

In what is perhaps the most provocative section of his treatise, Giles
attempts to show that Averroes is committed to attributing the power of
understanding to this man. Here Giles is almost certainly responding to the
set of Quaestiones de anima mentioned above, quite probably the work of
Boethius of Dacia. 31 Even though their author is much more interested in the

31) Cf. Giele in Trois Commentaries, especially Liber secundus , q. 4, “Utrum anima intellectiva
sit secundum suam substantiam coniuncta corpori sicut perfectio eius substantialis”, pp. 68-77.
For details on the attribution of this text to Boethius of Dacia, see the editor s introduction, espe¬
cially pp. 15-20. Van Steenberghen (op. cit., p. 117) labels the defenders of the proposition “radical
Averroists”, as opposed to the “moderate Averroists”. Siger of Brabant, who nowhere defends the
proposition, appears to belong to the latter faction. The second problem of Albertus Magnus’ De
XVproblematibus (Opera Omnia 17, p. I, ed. Bernhard Geyer [Aschendorff, 1975], pr. II, p. 34)

B. F. Conolly / Vivarium 45 (2007) 69-92

separateness of the possible intellect than in its uniqueness , 32 these Quaestiones
are striking for the bold and clever defense of Averroes that they offer. For the
author attempts to undermine in one stroke those arguments against Averroes
that maintain that the Commentator cannot account for how this man under¬
stands, by affirming and then defending precisely what the arguments of Aqui¬
nas and others assert against Averroes, namely, homo non intelligit , that this
man does not understand. Indeed, the author points out that the critics of the
Commentator only assume it to be true that homo intelligit , that is, that this
man understands; nowhere do they prove it.

Several arguments are put forward by the author for this, including: that
understanding has no organ and so a man does not understand as he senses;
that this position is consistent with what Aristotle says; that supercelestial
bodies understand, and yet the intelligences are related to them not as form to
matter, but as mover to thing moved. 33 The most interesting and significant
argument, however, concerns how the appearance that I experience myself to
understand is explained away. The author asserts in reply that it is not /, but
the separated intellect that experiences this. 34 Giles’ charitable account of the
argument further explains that what / am really experiencing in such cases is
simply that the intellect is really united with me.

And if you say instead that a man u nderstands because each experiences in himself that he
understands, they respond <by saying> it is not true that I would experience myself to
understand, in such a way that the act of understanding would be attributed to me. Rather,
I experience myself to understand insofar as it is true that I experience the intellect to be
really united to me. 35

also appears to be a response to these radical Averroists, though there is not such clear evidence as
there is in Giles’ treatise (see below) that these Quaestiones de anima were the target of his attacks.

32) Even if the separateness of the possible intellect implies its uniqueness, the author does not
explicitly raise the latter issue. In this respect, the particular quaestio in which these issues are
discussed is much more an attack upon Aquinas’ theory of intellect than they are a defense of Aver-
roism. For the author argues at length against the two cornerstones of Aquinas’ theory of intel¬
lect: (1) that the intellective soul is related to the body as form to matter (the author preferring
instead the other alternative offered by Aristotle, namely that they are related as captain to ship),
and (2) that the intellect is separated from the body in operation but not in substance. An attack
on Aquinas is not necessarily a defense of Averroism.

33) On these arguments, cf. Giele, in Trois Commentaires , pp. 75-76.

34) The argument is quoted and discussed below.

35) De plurificatione , 58r C: “Et si tu dicis immo homo intelligit, quia unusquisque in seipso
experitur, quod intelligit, respondent non est verum, quod ego experiar me intelligere, ita quod
actus intelligendi mihi conveniat, sed experior me intelligere, pro tanto verum est, quod experior
intellectus esse mihi realiter unitum.”

B. F. Conolly / Vivarium 45 (2007) 69-92

On the basis of such arguments the author of the anonymous Quaestiones
maintains that it is therefore an abuse of the term to say that I understand ; and
furthermore, those arguments showing that the power of understanding can¬
not be attributed to the particular man, far from refuting the separateness of
the intellect, merely confirm the theory. Averroes therefore appears to be vin¬
dicated to that extent.

Giles offers three responses to these Averroistic arguments. The first
addresses the problem in itself, the second is purely polemical, 36 and in his
third response, Giles tries to show that the defense of homo non intelligit in fact
does not preserve the opinion of the Commentator. I shall address the last
argument first; that is, I shall examine whether Averroes does indeed attribute
the power of understanding to the individual man.

2.1. Homo non intelligit Is Not the Opinion of the Commentator

Giles argues that Averroes is indeed committed to maintaining that the power
of understanding is properly attributed to this man.

For as from the position of the Commentator it seems more to follow that it is more proper
that a man understands than that the intellect understands. Nor was the Commentator of
the position that a man does not understand, if anyone would diligently investigate what
he said. They argue that because the intellect is not the power of an organ, as sense <is>,
and so, though it is proper that a man senses, it is nevertheless not proper that a man
understands. It is apparent enough, however, that this argument does not have a stable
foundation. For though the intellect is not the perfection of an organ, it is nonetheless the
perfection of the essence of the soul. And because the essence of the soul is the act of the body,
it follows from this that this man understands; and because the essence belongs to him, it
is his power; and because the power belongs to him, it is his act. For the essence and the
power of the soul is ours. Therefore, understanding is ours; hence, we understand. 37

36) De plurificatione , 58r F. Giles’ second response argues that, since the author admits that he
does not understand, he is no different than a plant in this regard and no one should bother even
trying to speak to him. The author is aware of this objection, mentions it explicitly (cf. Giele in
Trois Commentairesy 75), and rightly passes over it.

37) De plurificatione , 58r F-G: “Quia ut ex positione commentatoris videtur sequi magis quod
magis sit propria ilia homo intelligit quam ilia intellectus intelligit. Nec commentator huius
positionis fuit quod homo non intelligat si quis diligenter sua dicta perquirat, et quod ipsi
arguunt quod intellectus non est virtus organica sicut sensu et ideo licet sit ista propria homo
sentit non tamen est ilia propria homo intelligit, satis apparet quod ratio non habet stabile
fundamentum. nam licet intellectus non sit perfectio organi est tamen perfectio essentie anime
et quia ipsa essentia anime est actus corporis, ex hoc sequitur quod ipse homo intelligat et quia
cuius est ipsa essentia, eius est potentia et cuius est potentia eius est actus, quia essentia et poten-
tia animae est nostra igitur intelligere est nostrum ergo nos intelligimus.”

B. F. Conolly / Vivarium 45 (2007) 69-92

The argument, as an explanation of the Commentators position, is essentially
correct. 38 Averroes does indeed appear to attribute the power of understanding
to the individual human being. However, that Giles should preserve the posi¬
tion of the Commentator raises difficult problems with regard to both Aver¬
roes’ doctrine itself as well as Giles’ argument against Averroes. 39 For the whole
point of Aquinas’ and Giles’ argument against Averroes has been to show that
Averroes cannot account for the communication of the power of understand¬
ing from the separated possible intellect to the individual human being. In
other words, that strategy assumes that the power of understanding is attrib¬
uted to this man only insofar as the separated possible intellect is conjoined to
the imaginative faculty of this man by means of the intelligible species. Yet
here Giles is claiming that on Averroes’ theory the power of understanding
must be attributed to the individual human being because the intellect is the
perfection of the essence of the soul, and the essence of the soul is ours. 40 In
other words, Giles is here effectively attributing to Averroes reasons for which
the power of understanding is to be attributed to this man different from the
reasons he and Aquinas have argued against. Giles does not appear to offer
arguments against the position he is here attributing to Averroes.

In any case, that Giles has argued that Averroes is indeed committed to the
notion that the power of understanding must be attributed to the individual
human being, is a significant departure from Aquinas. For Aquinas argues
against Averroes, not ad hominem or ad positionem , but ad rem. That is, he
shows that Averroes’ position is contrary to what Aristotle teaches and the
truth itself insofar as Averroes’ doctrine leads to the absurd consequence homo

38) It is correct even if somewhat simplified. Averroes’ theory is actually quite complex and
difficult on this point, and can be developed only by a thorough exposition of Averroes’ theory
of conjunction, as discussed in the difficult and obscure text 36 of the Magnum Commentarium.
The context here (“from the position of the Commentator...”) suggests that Giles is drawing out
the consequence from Averroes’ position rather than appealing to something that the Com¬
mentator expressly says. Giles could have referred to several places, however, especially texts 3
(p. 390), 18 (pp. 439-440), and 36 (pp. 493-497) of the Magnum Commentarium , where Aver¬
roes makes it quite clear that the power of understanding is to be attributed to the individual
man, since abstraction and understanding “are led back to our will”.

39) I shall leave for another occasion the complex difficulties this raises for Averroes’ doctrine.

40) In any case, it is not clear that Giles really believes that the position of the Commentator is
that a man understands. For in both earlier and later treatises he appears to maintain expressly
the opposite. Cf. Giles of Rome, Errores philosophorum , ed. Josef Koch, trans. John O. Riedl
(Milwaukee, 1944), 22-25, esp. n. 57, and Quodlibet II, q. 20, (Louvain, 1646; repr. Frankfurt/
Main, 1966), respectively.

B. F. Conolly / Vivarium 45 (2007) 69-92

non intelligit. Aquinas nowhere tries to show that Averroes is otherwise and
elsewhere committed to attributing the power of understanding to the indi¬
vidual human being, and is thus internally inconsistent. This is the reason
why Aquinas Averroist critic could boldly affirm homo non intelligit against
Aquinas and in defense of Averroes.

Giles, by contrast, argues first ad hominem and only afterwards argues ad
rem. That is, he first shows that Averroes 5 position is internally inconsistent.
For it leads to the consequence homo non intelligit , which is inconsistent with
what Averroes maintains otherwise (thus “bracketing 55 for the moment the
question whether homo non intelligit is really absurd in itself). It is only after
he has established this that Giles argues ad rem , that is, only then does he go
on to show that homo non intelligit really is an absurd consequence of Averroes 5
doctrine. Giles 5 argument therefore proceeds against the Commentator in a
much more dialectically effective manner than does Aquinas 5 .

2.2. Homo non intelligit Considered Independently of the Opinion of Averroes

The problem of whether this man understands must still be considered on its
own terms, that is, independently of whether or not it is the opinion of Aver¬
roes. It is in his first response to the Latin Averroists who affirm that homo non
intelligit that Giles tries to show that whether they want to or not, Averroes
and the Latin Averroists must accept the position that the intellective power
must be communicated to this man. For Giles maintains that the Averroist
contradicts himself insofar as he concedes that a man does not understand, yet
says that he perceives the understood thing to be naturally conjoined to him.

For they concede that a man does not understand and yet they say that they perceive the
understood thing to be naturally conjoined with them. But perceiving the understood
thing to be naturally conjoined with them can happen only through the act of the intellect.
For the act of sense does not extend itself to perceiving the understood thing or the intel¬
ligible thing. Therefore by perceiving this we understand; but earlier they conceded that we
do not understand; therefore, they contradict themselves, since according to them we
understand and we do not understand. But if they still wanted to defend their position,
<then> instead of the incongruence, they would have to deny what they earlier conceded,
namely, that we perceive the understood thing to be naturally united with us. For it must
be posited that reason forced them to this. For, if they say this, <then> either they perceive
this and so contradict themselves, or they are speaking fantastically and then they are not
to be listened to, as neither are the authors of fables. 41

41) De plurificatione , 58r E-F: “Nam isti concedunt quod homo non intelligit et tamen dicunt
quod percipiunt intellectum eis esse naturaliter coniunctum, sed percipere intellectum eis esse

B. F. Conolly / Vivarium 45 (2007) 69-92

It is difficult to assess Giles’ analysis of the argument as a response to what we
actually find in the Averroist texts available to us from this period. This is
because the position that Giles argues against is not exactly what we find in
those texts. Consider, for instance, the following passage from the Quaestiones
de Anima mentioned above:

You say: I experience and perceive myself to understand; I say that this is false. Rather the
intellect united to you naturally, as a mover and ruler of your body, that is who experiences
this, just as the separated intellect also experiences things that are understood in them¬
selves. If you say, I, the aggregate of body and intellect, experience myself to understand, it
is false. Rather, the intellect needing your body as object experiences this, communicating
this to the aggregate in the said way. 42

Nowhere do we find this Averroist saying that “I”—understood as the aggre¬
gate of body and intellect—perceive myself to understand, contrary to Giles’
account of their position. On the contrary, the anonymous author expressly
denies this to be true. Indeed, he has no choice. For, as Giles correctly implies
in his argument it is only through a comprehensive cognition—i.e., under¬
standing or intelligens —that understanding could be known as understand¬
ing. That is, strictly speaking, one simply cannot perceive oneself to be
understanding; rather, one can only understand oneself to be understanding.

Giles’ argument therefore has no effect upon this Averroist s position: there
is no contradiction if one simply denies, as our Averroist has done, that I, the
aggregate of body and intellect perceive or otherwise experience myself to
understand. Neither is the Averroist’s position absurd in itself, since the Aver¬
roist does not deny that there is this reflective experience in which understand¬
ing understands itself. Rather, the question that remains is who is the subject
of this experience? For surely there is a reflective cognition of understanding

coniunctum non potest esse nisi per actum intelligendi quia ad percipendum intellectum vel
intelligibile non se extendit actus sensus igitur hoc percipiendo intelligimus ergo sibi contra-
dicunt cum secundum eos intelligamus et non intelligamus quod si tamen suam defendere vel-
lent positionem non haberent pro inconvenienti negare quod prius concesserunt scilicet quod
percipimus intellectum esse nobis naturaliter unitum cum ad hoc ponendum ipsa ratio eos
cogat, quia si hoc dicunt vel hoc percipiunt et tunc sibi contradicunt vel phantasmatice loquun-
tur et tunc non sunt audiendi ut nec conditor ipse fabularum.”

42) Anonymous, in Giele, Trois Commentaires , p. 76: “Tu dices: ego experior et percipio me intel-
ligere. Dico quod faisum est. Immo, intellectus unitus tibi naturaliter, sicut motor tui corporis
et regulans, ipse est qui hoc experitur sicut et intellectus separatus experitur intellecta in se esse.
Si dicas: ego aggregatum ex corpore et intellectu experior me intelligere, faisum est. Immo intel¬
lectus egens tuo corpore ut obiecto experitur hoc, communicans illud aggregato dicto modo.”

B. F Conolly / Vivarium 45 (2007) 69-92

that identifies the “I” who is the subject of understanding with the “I” who is
the subject of sense. Such experience cannot simply be dismissed, and part of
that experience reveals that when I understand myself to be the subject of
understanding, I acknowledge a unity of experience that comprehends my
sense-experience as well. In assessing the responses to this decisive question, it
must be recalled that the Averroists are working from a fundamentally different
conception of the relation between intellect and body than is either Giles or
Aquinas, and this difference will determine the answer to the question of who
is the subject of this unity of reflective experience. Giles and Aquinas under¬
stand the soul to be the form of the body, and the intellect to be a power of
the soul. The intellective soul is thus united to the body in such a way that the
composite of body and soul constitute a single subject —this human being —
which is the subject of all the soul’s powers, including intellect. The Averroists,
by contrast, see the relation of intellect to body in quite a different way. For
them, intellect is united to body not as form to matter but as mover to instru¬
ment or thing moved. This is emphasized repeatedly by the author of Quaes -
Hones de Anima. On this theory, there is not one subject constituted by the
union of intellect and body, but two subjects, namely, the body (as informed
by the soul), which is subject to all the soul’s powers (apart from understand¬
ing), and intellect, which is the subject of the power of understanding.

It may be, then, that on the Averroist theory, it is the separated intellect
who, in experiencing itself to understand, also recognizes a unity of experience
that comprehends the sense experience of this man as well. Consequently,
without addressing and attempting to undermine this Averroist conception of
the relation between body and intellect, or at least, without arguing that the
Averroists cannot account for the reflexive cognition of understanding that
identifies the “I” who is the subject of understanding with the “I” who is the
subject of sense, Giles’ refutation of the Averroist doctrine of the unique and
separated intellect appears not to be complete.

BRILL

Vivarium 45 (2007) 93-112

VIVA

RIUM

www.brill.nl/viv

La structure de l’acte intellectif dans les
theories ockhamiennes du concept

Ernesto Perini-Santos

Departamento de Filosofia, Universidade Federal de Minas Gerais ,
Belo Horizonte , Bresil

Abstract

William of Ockham held in his career two different theories about the nature of con¬
cepts. According to the first theory, concepts are forged by the mind and “terminate”
the mental acts which produce them. This so called “fictum”-theory was abandoned,
and Ockham held another theory, according to which concepts are identified with the
mental acts themselves. While I think this is a correct description of the evolution of
his philosophy, there is one aspect that has gone so far (almost) unnoticed : in his later
theory, not only concepts do not terminate mental acts, but nothing seems fit to play
this role. Mental acts are no longer “terminated” by anything. Therefore, as the theory
of concepts changes, there is also a change in the theory of mental acts. This last
change explains the disappearance of the vocabulary associated with the verb “termin-
are” in the exposition of the mental act theory.

Keywords

William of Ockham, concepts, cognition, mental acts«

II est bien connu qu’au long de sa carriere philosophique, entre 1317, date de
son commentaire aux Sentences fait a Oxford, et 1325, annee de la redaction
finale des Quodlibeta a Avignon, Guillaume d’Ockham a soutenu differentes
theories sur la nature du concept 1 . La premiere theorie, qui associe le concept

0 Pour une breve presentation de la vie et de la chronologie de l’oeuvre ockhamiennes, voir Paul
Vincent Spade, “William of Ockham”, dans: Stanford Encyclopedia of Philosophy, http://plato.
stanford.edu/entries/ockham. Cfr. aussi William J. Courtenay, “The Academic and Intellectual
Worlds of Ockham”, dans The Cambridge Companion to Ockham, ed. P. V. Spade (Cambridge,
1999), 17-30.

DOI: 10.1163/156853407X202520

E. Perini-Santos / Vivarium 45 (2007) 93-112

a une entite forgee par fame ayant un etre objectif, un fictum , a ete abandon-
nee au profit de l’identification du concept a facte meme d’intellection 2 . Une
telle evolution de la pensee ockhamienne a ete etablie deja par Philotheus
Boehner, et il ne semble pas quil faille revenir sur ce point. En fait, le passage
d’une theorie a l’autre est un critere pour la determination de la chronologie
des textes ockhamiens 3 .

En effet, dans ses premieres oeuvres, Ockham adhere pleinement a ce qu’il
est convenu d’appeler la theorie du fictum. Dans les textes de la Reportatio oil
il est question de la nature du concept, cest la seule alternative en vue 4 . Il en
va de meme dans une question disputee sur factivite de fintellect, contempo-
raine de la redaction de la Reportatio , qui offre une breve description de la
genese du concept universel. A partir de la connaissance d’un singulier causee

2) Cfr., par exemple, Philotheus Boehner, “The Relative Date of Ockhams Commentary on
the Sentences”, dans: Collected Articles on Ockham (St. Bonanventure, 1992), 96-110, surtout
99-107 ; Marilyn Adams, William Ockham (Notre Dame, 1987), 73-107 ; Gedeon Gal, “Gual-
teri de Chatton et Guillelmi de Ockham Controversia de Natura Conceptus Universalis”, Fran¬
ciscans Studies 27 (1967), 191-212 ; Elizabeth Karger, “Theories de la Pensee, de ses Objets et de
son Discours chez Guillaume d’Occam”, Dialogue 33 (1994) 437-436 et “William of Ockham,
Walter Chatton and Adam of Wodeham on the Objects of Knowledge and Belief”, Vivarium 33
(1995), 171-196 ; Francis Kelley, “Some Observations on the ‘Fictum’ Theory in Ockham and
its Relation to Hervaeus Natalis”, Franciscan Studies , 38 (1978), 260-282 ; Robert Pasnau, Theo¬
ries of Cognition in the later Middle Ages (Cambridge, 1997), 277-289 ; Claude Panaccio, Le
discours Interieur (Paris, 1999), 258-264.

3) “In favor of this relative chronology we have advanced two main reasons. The first is based on
a development of a certain doctrine from one opposite to the other. As is well known, Ockham
has advanced two theories concerning the nature of universal. The one is called the “ Fictum -
theory”. According to this a concept or a universal does not have real being, and therefore is not
an accident of the soul as its subject, but has only the being of an object of thought, an “esse
objectivum”, a purely ideal existence. This is called, following St. Augustine (and Abaelard),
“ fictum ” (not fiction !). The other theory is called “Intellectio-theory”, according to which the
concept or the universal is a real accident of the soul, identical with the act of understanding
(intellectio) and hence it exists in the soul as subject, it has an 11 esse subiectivum n .” (Boehner, “The
Relative Date of Ockhams Commentary on the Sentences”, 99-100.)

4) Il en est ainsi dans un passage oil Ockham critique la theorie scotiste de la species : “Item,
consequentia sua non valet: ‘intellectus potest habere obiectum praesens, igitur habet speciem
intelligibilem’. Quia aut intelligit per obiectum praesens aliquod fictum et ens rationis, aut ens
reale. Si primo modo, tunc non oportet propter tale obiectum ponere aliquam speciem praeviam
cognitioni repraesentantem, quia illud ens rationis causatur per actum intelligendi”, Quaestiones
in Librum Secundum Sententiarum - Reportatio , ed. G. Gal et R. Wood (OTh V), II, q. 13,
p. 299, 18-23. Cf. aussi, dans cette meme question, p. 304, 24-305, 3 (dorenavant, ‘Rep. II’).
G. Gal et R. Wood, les editeurs de ce texte, rapportent en note que, dans un codex , a cote de ce
passage, le changement d’opinion d’Ockham sur ce point est indique.

E. Perini-Santos / Vivarium 45 (2007) 93-112

par l’objet dont elle est la connaissance, ou par une disposition acquise de
Tesprit ( habitus ) produite par un acte cause par l’objet, facte cognitif termine
dans un etre objectif, dans un fictum , qui est le produit de cet acte meme 5 6 7 .
Dans les textes plus tardifs, comme la Somme de Logique G ou les Quodlibetc ?’,
le concept est simplement identifie a un acte de Tesprit {actus intelligendi ).

La solution adoptee par le Venerabilis Inceptor est construite surtout dans
F Ordinatio et dans le commentaire du Peri hermeneias , on y trouve une longue
discussion sur la nature du concept, avec des arguments pour et contre ces
positions. En fait, il nest pas simple d’identifier la chronologie de la construc¬
tion de la theorie du concept comme acte dans ces deux textes. Les commen-
taires de XArs Vetus sont globalement situes apres YOrdinatio: les premiers
auraient ete faits a Londres, entre 1321 et 1324, alors que le commentaire des
Sentences a probablement ete compose a Oxford, entre 1317 et 1319 8 . II y a
toutefois au moins deux passages dans Y Ordinatio qui renvoyent a YExpositio
du Peri hermeneias 9 . S’il n’est pas sur qu’il y ait plus d’un strate textuel dans
f Ordinatio™ , il ne me semble pas necessaire de postuler un ordre chronolo-
gique pour la totalite des textes de ces deux commentaires. En outre, dans ces
deux textes, il nest pas du tout clair qu’Ockham epouse de fa^on definitive la

5) “Et habita notitia, statim ad eius praesentiam - si non sit impedimentum - sequitur natu-
raliter alius actus, distinctus a primo, terminatus ad aliquod tale esse obiectivum quale prius vidi
in esse subiectivo. Et ille actus secundus producit ilia universalia et intentiones secundae et non
praesupponit eas.”, Quaestiones Variae , ed. G. Etzkorn, F. E. Kelley et J. C. Wey (OTh VIII),
q. V, quaest. disp. 3, 406-410, p. 175 (dorenavant, ‘Quaest. Var.’). Sur le sens de ‘intention
seconde’ dans ce passage, cf. Karger, “Theories de la Pensee”, 440-441.

6) Cfr. Guillaume d’Ockham, Summa Logicae , ed. Ph. Boehner, G. Gal et S. Brown (OPh I), I,
12, 29-39, pp. 42-43 (dorenavant, ‘S.L.’).

7) Cfr. Guillaume d’Ockham, Quodlibeta Septem , ed. J. C. Wey (OTh IX), IV, q. 35, pp. 469-
474 (dorenavant, ‘Quod.’).

8) Guillaume d’Ockham, Expositio in Librum Perihermeneias Aristotelis , ed. A. Gambatese et
S. Brown (OPh II), Introductio, 14*-15* (dorenavant ‘Exp. Per.’) ; cfr. Paul Vincent Spade,
“William of Ockham”.

9) Guillaume d’Ockham, Scriptum in Librum Primum Sententiarum - Ordinatio , ed. G. Etzkorn
et F. Kelley (OTh IV), I, d. 27, q. 3, p. 242, 18-243, 5 (dorenavant ‘Ord. IV’). Cfr. aussi Scrip¬
tum in Librum Primum Sententiarum - Ordinatio , ed. S. Brown et G. Gal (OTh II), I, d. 2, q. 8,
p. 291, 7-15 (dorenavant ‘Ord. II’).

,0) Selon les editeurs de XOrdinatio, il y a des redactions plus ou mois completes, plutot qu’un
ordre chronologique entre les differents sources textuels (Ord., I, Introductio, 19-20*). Cfr.
toutefois le texte cite dans la note 31 ci-dessous, pour une insertion tardive dans le texte de
XOrdinatio qui suggere bien l’ajout de la theorie du concept comme acte a un texte qui ne prenait
en consideration que la theorie du fictum.

E. Perini-Santos / Vivarium 45 (2007) 93-112

theorie du concept comme acte, plutot qu’il ne l’expose “ recitative” u . Dans le
commentaire du Peri hermeneias la theorie du concept comme acte est presen¬
tee comme la plus probable, mais Pidentification du concept a un fictum ayant
un esse obiective semble a Ockham aussi une opinion plausible 12 .

J’examinerai ces deux textes comme etant d’une periode intermediaire,
pour ce qui concerne la theorie du concept. Ils presentent les discussions les
plus developpes de ce sujet, et si Ockham y penche vers la theorie du concept
comme acte, il ne prend pas une position aussi tranchee que dans ses textes
plus tardifs. Sur un point precis, que j’essaierai de mettre en lumiere, je crois
que I’on peut identifier une approche de la question mieux organisee dans
I’ Expositio du Peri hermeneias que dans P Ordinatio. Cela ne signifie pas pour
autant que Pon doive identifier ici un ordre chronologique, meme si, dans ce
cas, il serait compatible avec la datation en general acceptee de ces textes.
Encore fois, il me semble plus avise de traiter ces deux textes comme appar-
tenant a une meme periode, pour ce qui concerne revolution de la theorie du
concept 13 .

Quoi qu’il en soit de l’ordre des textes, cette evolution est bien acceptee
dans la litterature : Ockham se rend compte notamment quil est superflu de
postuler des ficta. Comme il est dit dans une question des Quodlibeta , tout ce
qui peut etre sauve en postulant des ficta peut l’etre avec la seule postulation
des actes 14 . Il y a toutefois un aspect de ce changement theorique sur lequel je
voudrais attirer Pattention et qui, me semble-t-il, a deux exception pres, n’a
pas ete note dans Pimportante litterature sur ce sujet: le changement dans la
structure de Pacte intellectif qui accompagne le passage d’une theorie du con¬
cept a Pautre. Dans la premiere theorie, le concept est ce qui termine un acte
intellectif, alors que dans la theorie du concept comme acte, rien ne semble
devoir jouer ce role. John Boler 15 et Claude Panaccio 16 ont montre, dans des
travaux recents, quil y a un changement dans la theorie du concept qui se joue

10 Voir le texte cite dans la note 9 et les remarques des editeurs de ce volume dans Ord. IV,
Introductio, 15*-18*.

12) Exp. Per., I, Prooe., 7, pp. 359-363.

13) Dans l’examen d’un autre sujet, il m’a semble que l’ordre chronologique propose par les
editeurs ne correspondait pas exactement a revolution de la theorie ockhamienne, en l’occurrence,
sur les modalites - cfr. Ernesto Perini-Santos, “L’Extension de la liste des modalites dans les com¬
mentaries du Perihermeneias et des Sophistici Elenchi de Guillaume d’Ockham”, Vivarium 40
(2002), 174-188. Ici, encore une fois, je me limiterai a l’exposition de revolution d’une theorie
ockhamienne, sans considerer les consequences pour la datation de ses oeuvres.

,4) Quod. IV, q. 35, 115-129, p. 474.

15) John Boler, “Ockham on the Concept”, Medieval Philosophy and Theology 11 (2003), 65-86.
,6) Claude Panaccio, Ockham on Concepts (Hampshire, 2004), 27.

E. Perini-Santos / Vivarium 45 (2007) 93-112

autour de l’abandon de la these selon laquelle un acte cognitif doit terminer
dans un objet quelconque. Ils ont raison sur ce point, mais la description qu’ils
offrent de cette transition theorique chez Ockham me semble soit trompeuse,
pour Boler, soit insuffisamment developpee, pour Panaccio.

Dans la question 8 de la deuxieme distinction du premier livre de P Ordinatio,
consacree a la nature du concept universel, la premiere opinion envisagee
identifie le concept a l’intellection 17 . C’est essentiellement la meme opinion
que I ’Expositio sur le Peri hermeneias presente comme la position la plus pro¬
bable 18 . Dans ces deux textes, une objection est immediatement soulevee con-
tre cette these, \10rdinatio en offre la formulation suivante :

Contre cette opinion, on peut objecter : comme par toute intellection, quelque chose est
apprehende {aliquid intelligitur), done par cette intellection quelque chose est apprehende ;
mais non quelque chose singulier hors de 1’arne, parce que pas plus une chose que l’autre,
pas plus ce qui nest pas que ce qui est; done par cette intellection, ou bien rien nest
apprehende, ou bien n’importe quelle [chose] ; mais [il n’est pas vrai que] n’importe quelle
chose [soit apprehendee], sinon un nombre infini de choses serait apprehende par cette
intellection ; done, rien n’est apprehende 19 .

Cette objection met en oeuvre le principe selon lequel a toute intellection,
comme a toute vision ou a tout acte de desir, selon l’extension de Targument
dans le commentaire du Peri hermeneias 20 , correspond un objet. S’il en est

17) “Prima opinio potest esse quod universale est conceptus mentis, et quod ille conceptus est reali-
ter ipsa intellectio, ita quod tunc universale non esset nisi intellectio confusa rei, quae intellectio,
quia ipsa non plus intelligitur unum singulare quam reliquum, ipsa esset indifferens et communis
ad omnis singularia.. ”, Guillaume d’Ockham, Ord. II, I, d. 2, q. 8, p. 267, 7-p. 268, 3 ; cfr. aussi
p. 289, 12-292, 2.

18) “Alia posset esse opinio, quod passio animae est ipse actus intelligendi. Et quia ista opinio
videtur mihi probabilior de omnibus opinionibus quae ponunt istas passiones esse subiective et
realiter in anima tamquam veras qualitates ipsius, ideo circa istam opinionem primo ponam
modum ponendi probabiliorem.. ”, Exp. Per., I, Prooe., 6, 4-8, p. 351.

,9) “Contra istam opinionem potest argui: quia omni intellectione aliquid intelligitur, igitur tali
intellectione aliquid intelligitur; et non aliquid singulare extra animam, quia non plus unum quam
alterum, nec plus illud quod non est quam illud quod est; igitur vel nihil tale vel quodlibet tale
intelligitur ilia intellectione ; et non quidlibet, | § quia tunc infinita intelligentur ilia intel lectione ;
§ | igitur nihil.”, Ord. II, I, d. 2, q. 8, p. 268, 7-13. Cfr. Exp. Per., I, Prooe., 6, 34-48, pp. 352-353.
20) Exp. Per., I, Prooe., 6, 33-40, pp. 352-353.

E. Perini-Santos / Vivarium 45 (2007) 93-112

ainsi, quel est l’objet d’un concept universel ? Ce ne peut etre un objet univer-
sel, il n’y a rien de tel dans l’ontologie ockhamienne ; ce ne peut etre un con¬
cept forge par l’intellect, c’est precisement la these a laquelle cette position est
une alternative. Ockham a recours a un argument par parite de raison assez
repandu au XIV* me siecle : ce ne peut pas etre une chose hors de Tame, puisque
comme il n’y a pas plus de raison qu’une chose plutot qu’une autre soit
apprehendee par un concept F, ou bien toutes les choses qui sont F le sont, ou
bien aucune ne Test.

Si Ton ne peut accepter qu’aucun individu ne soit apprehende, il est moins
clair ce qu’il faut penser de l’autre consequence : pourquoi une intellection ne
pourrait-elle pas etre l’apprehension d’un nombre infini ou indetermine
d’objets ? Le commentaire du Peri hermeneias presente un argument assez
proche de celui-ci et, d’une certaine fa^on, repond a cette question 21 . Par une
connaissance commune ( cognitio communis ), ou bien quelque chose est intel¬
lige, ou bien rien n’est intellige {autaliquidintelligitur ista cognitione aut nihil).
Il n’est pas vrai que rien ne soit intellige, done quelque chose est intellige. Ce
qui est intellige n’est pas une res universalis quia nulla talis est. Ce n’est pas plus
une chose singuliere qu’une autre, done ou bien toute chose singuliere est
intelligee, ou bien aucune ne Test; comme la derniere reponse est evidemment
fausse, toute chose singuliere est intelligee, e’est-a-dire, tous les hommes sont
intelliges par la cognitio qui correspond a ‘homme’, tous les animaux par celle
qui correspond a animal’, et ainsi de suite. Or, cela, dit XExpositio , ne semble
pas devoir etre accepte {quod videtur inconveniens).

Ockham repond a cette critique dans la suite du texte. On peut dire que

... par une telle intellection confuse sont apprehendees des choses singulieres hors de 1’arne,
ainsi par exemple avoir une intellection confuse de l’homme n’est pas autre chose qu avoir
une connaissance par laquelle un homme n’est pas plus apprehende qu’un autre, et toutefois
[dire] que par une telle connaissance un homme est plus connu ou apprehende qu’un ane.
Et cela n’est pas autre chose [que dire] qu’une telle intellection, par une certaine fa^on
d’assimilation, s’assimile plus a un homme qu’a un ane, et pas plus a cet homme qua cet
autre. Et selon cela, il semble que Ton doive dire en consequence que par une telle intellec¬
tion confuse, un nombre infini de choses peut etre connu 22 .

2,) Exp. Per., I, Prooe., 6, 33-48, pp. 332-333.

22) “... tali intellectione confusa intelliguntur res singulares extra, sicut habere intellectionem
hominis confusam non est aliud quam habere unam cognitionem qua non magis intelligitur
unus homo quam alius, et tamen quod tali cognitione magis cogniscitur sive intelligitur homo
quam asinus. Et hoc non est aliud quam quod talis cognitio aliquo modo assimilationis magis
assimilatur homini quam asino, et non magis isti homini quam illi. Et secundum illud videtur

E. Perini-Santos / Vivarium 45 (2007) 93-112

De meme que l’on peut desirer l’infinite de parties d’une quantite continue,
par cette intellection on peut apprehender une infinite d’individus.

Et ainsi on peut dire que la meme intellection peut etre d’un nombre infini de choses, ce ne
sera toutefois pas une connaissance propre a une d’entre elles, pas plus qu’une chose ne peut
etre distinguee d’une autre par cette connaissance, et cela en raison d’une certaine ressem-
blance speciale de cette connaissance a certains individus, et non pas a d’autres 23 .

Le Venerabilis Inceptor decrit l’intellection de l’universel a peu pres de la meme
fa^on qu’il l’avait fait dans la critique. En fait, rien ne s oppose a l’apprehension
d’un nombre infini de choses par un concept. II accepte simplement la con¬
sequence d’une intellection d’un nombre infini ou indetermine d’individus, ce
qui etait precisement juge faux dans XOrdinatio.

La suite du texte de X Ordinatio presente un autre argument contre
1’identification du concept a une intellection. Un concept, dit ce texte, est ce
qui termine un acte d’intellection ( omnes vocatur conceptus mentis quod termi-
nat actum intelligendi) . Dans le cas d’un concept singulier, l’intellection d’un
individu ne se termine pas en soi-meme, mais dans l’individu dont elle est
l’intellection. Or, si l’intellection d’un universel se terminait en soi-meme, par
le meme raisonnement, il faudrait dire que l’intellection d’un individu se ter¬
mine aussi en soi-meme, ce qui, selon Ockham, est faux. Done aucune intel¬
lection ne se termine en soi-meme 24 . Selon cette objection, l’intellection doit
se terminer dans un objet autre qu’elle-meme, done il y a necessairement une
autre chose, differente de l’acte intellectif, qui joue ce role, le conceptus mentis.
Si l’idee que seule une chose differente de l’acte peut le terminer est presente
dans XExpositio du Peri hermeneias , cette objection ne s’y trouve pas.

consequenter dicendum quod tali cognitione confusa possunt infinita cognosci.”, Exp. Per., I,
Prooe., 6, 89-97, p. 355. La meme reponse se trouve dans Guillaume d’Ockham, Quaestiones in
Libros Physicorum Aristotelis (OPh VI), ed. S. Brown, q. 7, 53-69, p. 408 (dorenavant ‘Quaest.
Phys.’).

23) “Sic igitur posset dici quod eadem cognitio potest esse infinitorum, non tamen erit cognitio
propria alicui illorum, nec ista cognitione potest unum distingui ab alio, et hoc propter aliquam
similitudinem specialem istius cognitionem ad individua ilia et non alia.”, Exp. Per., I, Prooe., 6,
105-109, p. 355.

24) “Praeterea, hoc secundum omnes vocatur conceptus mentis quod terminat actum intelli¬
gendi ; sed talis intellectio non terminat se ipsam primo, quia non est maior ratio quod una
intellectio terminet se ipsam quam alia ; et ita cum intellectio Sortis non terminet se ipsam primo,
igitur nec ista intellectio terminat se ipsam primo ; igitur conceptus non est ipsa intellectio.”,
Ord. II, I, d. 2, q. 8, p. 268, 14-19.

100

E. Perini-Santos / Vivarium 45 (2007) 93-112

La disparition de cette objection dans le commentaire du Peri hermeneias
semble indiquer que l’idee qu’un concept est ce qui termine un acte intellectif
perd sa force d’un texte a l’autre. Si cela est vrai, alors la theorie du concept
comme acte est plus solidement etablie dans le commentaire du Peri hermene¬
ias que dans celui des Sentences. Quoi qu’il en soit de fordre des textes, il
importe surtout de remarquer que dans la deuxieme theorie ockhamienne,
rien ne semble devoir terminer un acte intellectif, et done la premisse qui
ouvre la deuxieme objection a cette theorie dans YOrdinatio (omnes vocatur
conceptus mentis quod terminat actum intelligendi ) n’a plus aucune force. En
effet, l’expression ‘terminate n’apparait dans fexposition positive de la theorie
du concept comme acte ni dans YExposition , ni dans Y Ordinatio 1G .

Dans la partie de YExpositio consacree a l’examen de la nature du concept,
le verbe ‘ terminate apparait quatre fois. La deuxieme occurrence a lieu dans le
cadre de Imposition de la theorie du concept comme fictum : Yintentio animae
ayant un etre intentionnel termine facte d’intellection 27 . Les trois autres occur¬
rences postulent que facte mental se termine dans une qualite reellement
existante dans fame, differente de facte meme d’intellection. II en est ainsi
dans fintroduction de cette hypothese : le concept peut etre une qualite fame
reellement distincte de facte intellection, “terminant facte meme d’intellection
en tant qu’objet” ( terminans sicut obiectum ipsum actum intelligendi ) 28 . Dans
une serie de critiques contre la theorie du fictum , Ockham dit qu’une qualite
existant dans fame terminant facte intellectif peut, plus proprement que les
ficta , jouer les roles qui sont attribues a ces entites ayant seulement un esse
intentionale™ . La derniere occurrence, apres le refus de f attribution aux con¬
cepts d’un etre intentionnel, hors des categories, presente une autre solution :
les passions de fame sont des qualites mentales reelles, qui peuvent etre

... ou bien des intellections ( intellectiones ), selon une opinion, ou bien certaines qualites de
Tame terminant en tant qu’objects les intellections de 1’ame (terminantes tamquam obiecta
intellectiones animae ) 30 .

25) Exp. Per., 1,6, pp. 351-358 et 9, pp. 363-369.

26) Ord. II, I, d. 2, q. 8, p. 289, 12- p. 292, 2.

27) Exp. Per., I, 7,1. 23-24, p. 360.

28) Exp. Per., I, 4,1. 5-6, p. 349.

29) “Similiter, quamvis praedicta propter alias rationes non possunt competere intellectioni, qui
difficile est salvare quid intelligam tali intellectione, tamen omnia ilia possunt verius competere
alicui qualitati existenti in anima, quae terminat actum intelligendi..Exp. Per., I, Prooe., 7,
46-49, p. 361.

30) “... vel quia sunt intellectiones, secundum unam opinioem, vel quaedam qualitates animam
terminantes tamquam obiecta intellectiones animae.”, Exp. Per., I, 9,1. 163-165, p. 369. II me

E. Perini-Santos / Vivarium 45 (2007) 93-112

101

L’alternative n’oppose pas deux types d’objets qui termineraient l’acte
d’intellection, mas deux possibilites pour la theorie du concept, dont seule la
deuxieme correspond a quelque chose qui termine l’acte d’intellection 31 . Si
l’hypothese de l’identification du concept a une qualite reellement existante
dans Tame ne semble pas avoir ete acceptee par Ockham, elle confirme une
exigence qui figure aussi bien dans 1* Ordinatio - a une exception pres - que
dans I ’Expositio : seule une chose singuliere differente de l’acte intellectif peut
terminer celui-ci 32 .

Le vocabulaire associe a l’expression ‘terminare nest en effet pas complete-
ment absent des textes dans lesquels le concept est identifie a un acte
d’intellection. II y a un texte - le seul, a ma connaissance - dans X Ordinatio oil
un acte d’intellection est dit terminer dans les choses memes qu’il signifie :

On peut toutefois dire avec probability qu’il n’y a pas un tel intermediaire [,i.e. un fictum],
mais que 1’universel est alors la connaissance confuse elle-meme terminee immediatement
dans les choses singulieres auxquelles elle est commune et universelle... . 33

II me semble que le verbe ‘terminare ici est equivalent a signifier’. Si tel est le
cas, ce passage est certes un contre-exemple a ma these selon laquelle cette
expression disparait des tous les textes dans lesquels la theorie du concept

semble que la clause disjonctive dans Exp. Per., I, Prooe., 9, 165-169, p. 369, qui suit ce passage,
ne renvoie pas aux lignes indiquees par les editeurs, mais plutot a Exp. Per., I, Proee., 6, 110-145,
pp. 355-356.

31 > Cette meme disjonction apparait dans une insertion tardive dans le prologue de V Ordinatio.
II s’agit de determiner le sujet de la theologie, c’est-a-dire, le sujet des propositions theologiques,
ou plus precisement des conclusions theologiques. II faut distinguer deux cas, la theologie pro
statu isto et celle des bienheureux. Dans le premier cas, Dieu nest pas le sujet de la theologie,
“... quia illud est terminus conclusionis quod immediate terminat actum intelligendi | § vel est
actus intelligendi. § | Sed Deus in se non immediate terminat actum intelligendi sed mediante
aliquo conceptu sibi proprio, | § nec est conceptus. § | Igitur ille conceptus non est Deus, erit
subiectum theologiae nostrae.”, Guillaume d’Ockham, Ord. I, Prol., q. 9, p. 268, 24-269, 5.
Les signes ‘ | §’ et ‘§ | ’ marquent des ajouts a une version incomplete d’un codex florentin de
I’ Ordinatio (Ord. I, Introductio, 20*). II faut noter ici que ralternative n’oppose pas deux objets
dans lesquels se termine l’acte, mais deux termes d’une proposition, dont Tun est ce qui termine
1’acte d’intellection et l’autre est l’acte lui-meme.

32) “Posset igitur poni una talis opinio, scilicet quod passio animae, de qua Philosophus hie
loquitur, est aliqua qualitas animae distincta realiter ab actu intelligendi, terminans sicut obiec-
tum ipsum actum intellegendi, quae quidem qualitas non habet esse nisi quando est actus intel¬
ligendi.”, Exp. Per., I, Prooe., 4, 3-7, p. 349.

33) “Potest etiam dici probabiliter quod nihil est tale medium, sed quod tunc universale est
ipsamet cognitio confusa terminata immediate ad omnes res singulares quibus est communis et
universalis..Ord. IV, I, d. 27, q. 3, p. 242, 23- p. 243, 2.

102

E. Perini-Santos / Vivarium 45 (2007) 93-112

comme acte est acceptee, mais une exception anodine. Dans les deux cas
ou Ockham accepte clairement que le concept est ce qui termine facte
d’intellection (le concept comme fictum et le concept comme qualite existante
reellement dans Tame), il y a une distinction entre ce qui termine facte et ce
qu’il signifie. II est clair que le role du concept comme signe ne disparait pas
d’une theorie a f autre. II faut done dire que, selon ce passage, un concept
signifie et termine dans les choses singulieres; il nest pas tres clair l’utilite de
garder f expression ‘terminate et, surtout, quelle ait un role different de celui
de la signification. Outre le fait que les relata des relations indiquees par les
verbes ‘signifier’ et ‘terminer sont les memes, leur “direction” est la meme : le
concept est id ce qui demande un terme, a f inverse des deux autres theories
ockhamiennes dans lesquelles il etait le terme de facte.

Encore une fois, XOrdinatio semble avoir une version moins developpee de
la theorie du concept comme acte que XExpositio. Ce dernier texte isole ne me
semble pas presenter une veritable explication du concept ‘terminer dans le
cadre de la theorie du concept comme acte. Si Ton accepte que si le concept
termine dans les choses qu’il signifie, fidee meme d’un terme du concept, ou
du concept comme le terme d’un acte d’intellection, devient superfetatoire, la
quasi absence de cette expression dans f exposition positive de cette theorie
s’explique plus facilement. On peut en effet se demander qu’ est-ce qui pour-
rait terminer facte d’intellection dans la theorie du concept comme acte,
outres les choses singulieres ? Il me semble que la reponse est: rien. Mais si tel
est le cas, alors entre la theorie du concept comme fictum et celle du concept
comme acte, il y a un changement dans la theorie concernant la structure de
facte cognitif.

Un tel mouvement theorique n’est pourtant jamais explicite, a ma connais-
sance, dans les textes ockhamiens. Pour etayer un peu plus l’hypothese d’un
changement si important passe en sous-main, je propose de voir la structure
de facte intellectif que fon retrouve dans la theorie ockhamienne du concept
comme fictum chez d’autres auteurs, structure qui n’est pas celle de la theorie
du concept comme acte. Avant de le faire, neanmoins, je passerai en revue,
brievement, la vision de ce point par deux importants interpretes d’Ockham,
Claude Panaccio et John Boler.

Ill

Comme je l’ai note au debut de f article, la disparation de l’expression ‘ termi¬
nate dans la theorie du concept comme acte et les consequences theoriques
qui en suivent n’ont pas attire f attention des interpretes d’Ockham, jusqu’a

E. Perini-Santos / Vivarium 45 (2007) 93-112

103

tres peu de temps 34 . Claude Panaccio, dans son livre Ockham on Concepts , note
qu’Ockham abandonne l’idee que tout acte cognitif doit terminer en quelque
chose. Le Venerabilis Inceptor a vu que cette presupposition n’etait pas neces-
saire pour expliquer le role du concept comme signe 35 . Cette remarque me
semble tout a fait juste, en pla^ant correctement F abandon de la presupposi¬
tion que tout acte de pensee doit terminer en quelque chose dans le cadre de
revolution de la theorie ockhamienne du concept. Panaccio ne developpe
toutefois pas beaucoup ce point. En effet, il ne suffit pas de dire que Facte
meme d’intellection peut etre un signe dans le langage mental, parce le con¬
cept dans la theorie du fictum avait aussi le role d’etre le terme de Facte intel-
lectif. Quid de cet autre role dans la theorie du concept comme acte ?

John Boler a consacre tout un article a ce qu’il a appele le “ Terminator Prin¬
ciple', dans lequel il pretend precisement attirer Fattention sur la complexite
des transitions entre les theories ockhamiennes du concept. Selon lui, la raison
la plus forte pour Fadoption de la theorie du concept comme acte se trouve
dans l’idee que le role central joue par le concept est celui d’etre un terme dans
le langage mental 36 . L’effort de Boler est de montrer que cette transition
theorique chez Ockham peut etre mieux comprise si l’on met en relief le role
du concept comme le terme d’un acte intellectif, suivant un principe qu’il
appelle “ Terminator Principle" , et qui se trouve, selon lui, dans le commentaire
au Peri hermeneias : tout comme la vision exige un objet vu, toute pensee ( cog -
nitio ) exige un objet connu 37 . Lorsqu’il vient a adopter la theorie du concept
comme acte, Ockham accepte toujours ce principe, mais non pas l’idee que
Facte de pensee exige un concept comme son objet 38 .

34) J’ai traite de ce point dans ma these de doctoral, soutenue en 2001 ; voir Ernesto Perini-
Santos, Modalite et Evidence (Universite Francis Rabelais, Tours, 2001), 16-67.

35) Claude Panaccio, Ockham on Concepts , 27. Sur l’importance de Identification des concepts
a des termes dans le langage mental dans revolution de la theorie ockhamienne, voir aussi Claude
Panaccio, Les Mots, les Concepts et les Choses (Montreal-Paris, 1992), 69-164 et Le discours Interi-
eur , 253-278.

36) John Boler, “Ockham on the Concept”, 70. Comme nous 1’avons vu, il s’agit de la these
soutenue par Panaccio.

37) Exp. Per., I, Prooe., 6, 34-48, pp. 352-353. Le meme argument se trouve Ord. II, I, d. 2,
q. 8, p. 268, 7-13.

38) “If we assume, as I think we should, that, in his final theory, Ockham subscribes to the same
principle [i.e., the “Terminator Principle”,] and not just the same formulation, it casts a different
light on what changes in his position and what does not. Briefly put, it means that, although
Ockham comes to reject the idea that an act of thought must have a concept as its object
(i.e., that the concept is the “something cognized”), he holds throughout a strongly relational
account of cognition as oriented to some object.”, Boler, “Ockham on the concept”, 72.

104

E. Perini-Santos / Vivarium 45 (2007) 93-112

L’ argument de Boler nest pas tout a fait clair, en tout cas il nest pas clair en
tant qu’un argument pour une certaine interpretation d’Ockham : il n’indique
pas toujours le texte ockhamien qu’il a en vue, et il est en particulier difficile
de voir le deployment de son “ Terminator Principle ’ chez Ockham. On peut
noter, par exemple, que l’expression ‘ terminare ne figure pas dans l’exposition
positive de la theorie du concept comme acte dans le commentaire au Peri
hermeneias , texte auquel il renvoie. Il est neanmoins possible d’identifier la
reponse qu’il offre a deux questions : Pourquoi Ockham a-t-il adopte ce prin-
cipe ? Pourquoi a-t-il estime, lors de sa premiere theorie, que le concept etait
l’objet d’un acte de pensee ? Selon lui, l’idee ockhamienne est que les pensees
sont identifies par leurs objets ou signifies ; les concepts comme ficta sont
postules precisement pour identifier un acte de pensee, en en identifiant
son objet 39 .

Le role de ce principe apparait plus clairement si nous partons d’une analo¬
gic avec la peinture. Imaginons que quelqu’un fasse un portrait de la reine
d’Anglaterre. On peut dire, en montrant ce portrait, Vest la reine d’Anglaterre’.
Il est clair que cette fa$on de parler n’a aucun poids ontologique (“ does no
ontological heavy lifting ), mais sert plutot a designer de qui est le portrait en
question. Le recours a l’existence objective chez Ockham est une fa^on de
determiner ce qui fait d’un acte intellectif cet acte, et non pas un autre, par la
determination de ce dont il est une representation, suivant une exigence etablie
par le “ Terminator Principle ’ : une pensee est individuee par son objet. Ce
principe decoule de l’idee que toute representation est comme un portrait
{ u all representation is like portraiture”)*®. Si je comprends bien Boler, cela veut
dire que tout comme on identifie un acte de vision par ridentification de
l’objet vu, et Ton identifie un portrait en en determinant le modele, de meme
Tidentification d’un acte intellectif se fait par 1’identification de son objet. La
reponse ockhamienne a la premiere question est que ce principe sert a identifier
un acte intellectif, en en designant l’objet.

Pourquoi le concept est-il l’objet d’un acte intellectif? Si dans le cas d’un
acte intellectif qui vise un singulier cet objet n’est autre que le singulier lui-
meme, dans un acte intellectif relatif a un universel, comme il n’y a pas de
chose universelle, il faut postuler un ftctum qui soit 1’objet de l’acte, et ce

39) “Suppose the problem Ockham is dealing with could be put in this way: How should one
characterize or identify a thought ? That is, what makes a thought the thought it is ? Or what
makes a thought different from another ?... the identifying property of a concept is relational:
thoughts differ because of their different objects or significata. This is the claim of the Terminator
Principle.”, Boler, “Ockham on the concept”, 73.

40) Boler, “Ockham on the concept”, 77.

E. Perini-Santos / Vivarium 45 (2007) 93-112

105

fictum nest autre chose que le concept. Lorsqu il adopte la theorie du concept
comme acte, le Venerabilis Inceptor garde toujours le “ Terminator Principle ’,
mais ne pense plus que f objet qui sert a identifier facte intellectif soit un con¬
cept ; il suffit d’associer le concept aux choses signifiees elles-memes 41 . Cette
modification dans f explication de fintentionalite cfun acte intellectif est
accompagnee par une modification dans la structure meme de facte : au lieu
de la structure ternaire de la premiere theorie, le concept comme acte a une
structure dyadique 42 .

La modification dans la structure de facte intellectif me semble etre en effet
un point central dans f evolution de la theorie ockhamienne, une modification
en quelque sorte passee sous silence par Ockham lui-meme et, jusqu’a il y a
peu, par les commentateurs aussi. Lexplication de la theorie ockhamienne du
concept par le “ Terminator Principle ’ me semble etre, en revanche, bien moins
satisfaisante, pour au moins deux raisons. La premiere raison est que le verbe
‘terminare n’apparait pas dans toute fceuvre d’Ockham, mais, comme j’ai
essaye de montrer plus haut, a une exception pres, se trouve seulement dans les
textes dans lesquels il accepte, ne serait-ce que comme hypothese, la theorie du
concept comme quelque chose de different de facte intellectif lui-meme, que
ce soit un fictum ou une qualite de f esprit. Il n’y a pas de raison textuelle, me
semble-t-il, pour utiliser ce principe pour expliquer toutes les phases de
f oeuvre ockhamienne.

A cette raison textuelle s’ajoute une raison theorique plus importante :
f explication de fintentionalite du concept ne se fait pas par f identification de
fobjet qui termine facte intellectif, mais essentiellement par une theorie de la
similitude - theorie a laquelle Ockham semble bien adherer meme quand il
accepte la theorie du concept comme acte 43 . Un concept est le concept de ce

41) Boler, “Ockham on the concept”, 80-81.

42) Boler, “Ockham on the concept”, 81.

43) Sur ce point, voir Panaccio, Ockham on concepts , 119-143. Boler, bien entendu, sait qu’au
moins lors de sa premiere theorie, Ockham adopte une theorie du concept comme similitude
(Boler, “Ockham on the concept”, 71-72), mais je pense qu’il n’en voit pas les consequences pour
son argument. Je crois que cette conclusion doit etre affaiblie pour les concepts singuliers. Voir,
par exemple, le texte suivant de la Reportatio : “Igitur eodem modo est in proposito: quod licet
intentio vel species, si poneretur, aequaliter assimileretur multis individuis, tamen ex natura sua
determinat sibi quod ducat intellectum in cognitionem illius obiecti a quo partialiter causatur,
quia ita determinat sibi causari ab illo obiecto quod non potest causari ab aliquo alio. Et
ideo sic in eius cognitionem ducit quod non ducit in cognitionem alterius.”, Guillaume
d’Ockham, Quaestiones in Librum Secundum Sententiarum - Reportatio , ed. G. Gal et R. Wood
(OTh V), II, q. 13, 289,1. 1-7 ; cfr. aussi 279,1. 10-14. Rien ne depend toutefois de ce point, un
rapport causal n’etant pas couvert par le “ Terminator Principle”.

106

E. Perini-Santos / Vivarium 45 (2007) 93-112

dont il est la similitude, et ceci est suffisant pour identifier un concept, pour dire
en quoi il differe d’un autre concept. Bien entendu, ‘similitude’ est un concept
relationnel, mais il ne s’agit pas du tout de la relation qui est indique par le verbe
‘terminate chez Ockham. Si Boler a bien vu le changement dans la structure
de facte intellectif qui accompagne le changement dans la theorie du concept
ockhamien, il me semble quil n’en pas identifie correctement le ressort 44 .

L’inclusion de la theorie du concept dans une theorie generale de facte est plus
claire chez d’autres auteurs plus ou moins proches cf Ockham. Jean Duns Scot
refuse f identification du concept a ce dans quoi se termine facte intellectif par
f exclusion de l’intellection des actions productives, celles qui exigent un
terme 45 . Le point est encore plus explicite chez Gauthier Burley, qui expose
f argument suivant pour la postulation des idola AG :

Tout acte et toute composition a un certain terme, comme il est dit dans Physique V ; done
1’acte intellectif [a] aussi [un terme], sinon il ne serait ni fini, ni distinct 47 .

44) Un autre defaut dans Interpretation de Boler apparait dans l’importance quil attribue au
discours sur des entites non existantes pour le developpement de la theorie du concept comme
fictum. En effet, il me semble quil est faux de dire que “Ockham first adopted the fictum theory
in the context of our talk of non-existent things.”, Boler, “Ockham on the concept”, 74. Dans
les textes qu’il cite a la note 2 comme pertinents pour le debat (Boler, “Ockham on the concept”,
65), pas plus que dans d’autres textes dans lesquels il est question de la theorie du concept
comme fictum , Ockham ne s’occupe pas des discours sur des entites non existantes, il ne s’agit
pas pour lui d’expliquer “our talk of non-existent things”. Ainsi dans la question de XOrdinatio
consacree a la theorie du concept, les deux paragraphes dans lesquels il est question des entites
non existantes (Ord. II, I, d. 2, q. 8, p. 273, 1. 15-18 et p. 284, 1. 9-18) ne sont pas du tout
centraux dans l’argument ockhamien pour la postulation des ficta , et ne donnent en aucun cas le
contexte pour leur introduction.

45) Jean Duns Scotus, Ordinatio - Liber Primus, ed. Balic (vol. VI), I, d. 27, 54-56, p. 86.

46) Il attribue cette position a Grosseteste ; Brown, l’editeur de Burley, note qu’il n’a pas trouve
cet argument sous la plume de l’eveque de Lincoln ; Stephen Brown, “Walter Burleys Quaestio-
nes in Librum Perihermeneais\ Franciscan Studies 34 (1974), 210«; cfr. aussi Stephen Brown,
“Walter Burleys Middle Commentary on Aristotle’s Perihermeneais\ Franciscans Studies 33
(1973), 42-134, 53 n.

47) “Omnis actus et omnis comparatio habet aliquem terminum, ut dicitur V Physicorum ; igitur
et actus intellectus, aliter non esset finitus et distinctus.”, Burley dans: Stephen Brown, “Walter
Burley’s Quaestiones in Librum Perihermeneais\ 1.51, 210. Cfr. “Walter Burley’s Middle Com¬
mentary on Aristotle’s Perihermeneais \ 1.12, 53-54.

E. Perini-Santos / Vivarium 45 (2007) 93-112

107

II s’agit d’un aspect de la theorie generale de 1’acte, comme le montre le renvoi
a la Physique : tout changement va d’un terme a un autre, autrement il ne
serait ni fini, ni distinct 48 . Midolum est le terminus ad quem de l’intellection,
l’argument pour son introduction renvoie a une theorie generale de l’acte, ou
d’un type d’acte. Comme l’acte intellectif est interne, il est necessaire de pos-
tuler un terme lui-meme interne, c’est-a-dire Y idolum produit dans l’intellect
par l’intellect 49 . Le verbe ‘terminate s’applique a un acte d’intellection comme
une extension a ce type d’acte d’une structure commune a d’autres actes. Le
terme d’une intellection n’est pas autre chose que ce qui la finit comme acte,
sans quoi elle serait indeterminee. On peut penser au modele de la vision et
associer l’objet vu comme ce qui met en terme a la vision au fictum comme
ce qui met un terme a l’acte intellectif; sans l’objet, l’un et l’autre seraient
indetermines.

Ce raisonnement est refuse par Gauthier Burley. Son deuxieme argument
s’attaque a l’application a 1’intellection de cette these generale :

En outre, aucun philosophe n’a pense que, par une action qui reste a f interieur de f agent,
quelque chose reellement different de Taction soit produit dans l’agent, mais par une action
allant aux choses exterieures, et non pas par une action restant a finterieur de f agent,
quelque chose est realise outre faction 50 .

L’acte intellectif n’est pas le type d’acte qui exige un terme, parce qu’il est un
acte interne. L’important pour notre propos est de voir que la postulation d’un
idolum produit par l’intellect dans lequel se termine l’acte intellectif depend
d’une certaine caracterisation de celui-ci. Le maitre realiste oxonien conclut
que le concept n’est pas autre chose que l’acte d’intellection lui-meme 51 . Le
refus de cette vision de l’acte intellectif, explicite chez Burley, est implicite
dans revolution de la pensee ockhamienne.

Cette idee est presente aussi dans les critiques adressees a la theorie du
fictum par Gauthier Chatton, confrere d’Ockham au couvent franciscain de

48) Aristote, Pyhsique , V, 1, 224 a34-b2.

49) “Sed cum actus intellectus sit manens in intellectu necessario terminabitur ad aliquid intra et
illud terminativum actus est tale idolum repraesentans rem extra.”, Burley, dans : Stephen Brown,
“Walter Burleys Quaestiones in Librum Perihermeneais\ 1.51., 210.

50) “Praeterea, a nullo philosopho invenitur quod per actionem manentem in agente sit aliquid
productum in agente realiter differens ab ilia actione, sed per actionem transeuntem ad extra est
aliquid operatum praeter actionem et non per actionem manentem in agente.”, Burley, dans :
Stephen Brown, “Walter Burleys Quaestiones in Librum Perihermeneais\ 1.61, 211.

50 Burley suit ici Guillaume de Ware, comme f indique Claude Panaccio, Le discours interieur ,
195-196.

108

E. Perini-Santos / Vivarium 45 (2007) 93-112

Londres, dont le role dans Involution dans sa theorie du concept est en general
admis 52 . Ces critiques se fondent en partie sur le principe d’economie, et en
partie sur certaines consequences ontologiques indesirables des ficta , liees a la
separabilite entre l’acte et son objet 53 . II est remarquable que les arguments
epistemologiques n’y jouent pas un role important: si Chatton indique la
menace d’une regression in infinitum dans la representation 54 , il ne parle pas
de l’impossibilite de la connaissance, comme le fera Ockham 55 .

Le changement de la theorie de 1’acte qui accompagne l’abandon de la theo¬
rie du fictum apparait chez Chatton comme un argument supplementaire,
puisque motive independamment des critiques adressees directement au
fictum .

Quoi qu’il en soit de ton argument, que comprends-tu par ‘apprehender une chose’ ? Si ce
n’est rien d’autre que le fait que l’intellection est dans l’esprit ce a partir de quoi la chose
re^oit sa denomination de fa^on extrinseque et est dite apprehendee, de sorte que rien
d’autre n’est requis, je te le concede. Mais alors il est clair qu’il n’est pas necessaire de poser
un fictum quelconque. Si tu penses qu’il existe quelque chose qui termine 1’acte, par exem-
ple un fictum ou un etre reel, terme sans lequel il ne pourrait y avoir d’intellection, c’est une
idee fausse, parce que l’intellection <est> une qualite absolue et n’a pas besoin d’autre terme
que d’etre re^ue dans un intellect premier 56 .

Si dire qu’une chose est apprehendee revient simplement a dire qu’une intel¬
lection de cette chose est dans l’intellect, il n’est point besoin de postuler en
outre quelque chose d’autre forge par l’intellect; la premiere partie de la cri¬
tique reprend le souci d’economie ontologique qui meut l’ensemble de la
question. Ce principe n’est toutefois operatoire que si 1’entite a eliminer n’est

52) Gal, “Gualteri de Chatton et Guillelmi de Ockham Controversia de Natura Conceptus Uni¬
versalis” ; cfr. toutefois Panaccio, Le Discours Interieur , 262-263.

53) Gal, “Gualteri de Chatton et Guillelmi de Ockham Controversia de Natura Conceptus Uni¬
versalis”, 201-203.

54) Gal, “Gualteri de Chatton et Guillelmi de Ockham Controversia de Natura Conceptus Uni¬
versalis”, 203-204.

55) Cfr. Ockham, Quod., IV, q. 33, 84-91, p. 473.

56) “Quidquid sit de argumento tuo, quid intelligis per ‘rem intelligi’ ? Si nihil aliud nisi intel-
lectionem esse in mente a qua res extrinsece denominatur et dicitur intelligi, ita quod nihil plus
ibi requiritur, concedo tecum. Sed tunc non oportet ponere aliquod fictum, planum est. Si intel-
ligas quod aliquid sit ibi quod terminet actum, puta fictum vel ens reale sine quo terminante non
potest haberi intellectio, falsa imaginatio est, quia intellectio [est] absoluta qualitas et termino
non indiget aliter quam in primo intellectu acceptum est.”, Chatton dans : Gal, “Gualteri de
Chatton et Guillelmi de Ockham Controversia de Natura Conceptus Universalis”, 204.

E. Perini-Santos / Vivarium 45 (2007) 93-112

109

pas necessaire pour les besoins de la theorie. La deuxieme partie de la critique
va done attaquer precisement ce qui semble rendre necessaire la postulation
des ficta , un terme ad quem pour l’acte intellectif. Or cette exigence est, dit
Chatton, une falsa imagination X intellection n’exige pas un objet dans lequel
elle puisse se terminer, puisque e’est une qualite absolue, et non pas un type
d’acte qui demanderait un terme pour sa determination. Un tel terme nest pas
necessaire a la structure interne de Facte intellectif, il faut seulement quelque
chose a laquelle renvoie la qualite absolue quest l’intellection 57 .

Pierre d’Ailly permet aussi de voir comment, dans la theorie du concept
comme acte, l’intellection ne se termine en rien. Apres avoir repris presque
mot pour mot la theorie ockhamienne des notitiae intuitiva et abstractiva ™,
il ajoute, pour ce dernier cas, une distinction qui n’existe pas chez Ockham :

La troisieme distinction est que la connaissance abstractive qui nous est naturellement pos¬
sible est double. En effet, une [connaissance abstractive] est celle par laquelle la chose meme
est immediatement connue en elle-meme et [la connaissance] se termine en cette chose
meme objectivement, de sorte que rien de distinct de cette chose ne termine cette connais¬
sance. L’autre [connaissance abstractive] est celle par laquelle la chose meme nest pas con¬
nue en soi, mais dans un autre, et [la connaissance] ne se termine pas en cette chose
immediatement, mais en sa species ou en son image existant dans Tame 59 .

L’important id pour mon propos est l’alternative pour ce en quoi se termine
la connaissance abstractive : elle se termine soit dans quelque chose existant
seulement objectivement ( obiective ), soit dans la species de ce dont elle est la
notitia. Or aucun de ces deux objets n’est disponible pour la theorie finale
ockhamienne sur le statut du concept. La theorie mature ockhamienne ne

57) “Ad aliud, cum dicis ‘conceptus est id quod terminat actum’ : sed actum reflexum terminare
non est nisi voce cognosci; et sic dico quod res terminat et non fictum. Res terminat intellectio-
nem, id est denominatur extrinsece ab intellectione in mente, sic dicendo ‘res intellegitur’.”,
Chatton, dans : Gal, “Gualteri de Chatton et Guillelmi de Ockham Controversia de Natura
Conceptus Universalis”, 205.

58) Cfr. Pierre d’Ailly, dans : Ludger Kaczmarek, “‘Notitia’ bei Peter von Ailly, Sent. 1, q. 3.
Anmerkungen zu Quellen und Textgestalt”, dans Die Philosophie im 14. und 15. Jahrhundert , ed.
O. Pluta, Amsterdam, 1988, 406-414.

59) “Tertia distinctio est, quod duplex est notitia abstractiva nobis naturaliter possibilis. Nam
quaedam est, qua res ipsa immediate cognoscitur in se ipsa et ad ipsam obiective terminatur ita,
quod nihil aliud ab ea distinctum terminat illam notitiam. Alia est, qua res ipsa non in se cog¬
noscitur, sed in alio, nec ad ipsam rem immediate terminatur, sed ad eius speciem seu eius
imaginem in anima existentem.”, Pierre d’Ailly, dans Kaczmarek, “‘Notitia’ bei Peter von Ailly,
Sent. 1, q. 3. Anmerkungen zu Quellen und Textgestalt”, 414.

110

E. Perini-Santos / Vivarium 45 (2007) 93-112

contient rien qui puisse remplir cette fonction, quelque chose a la fois different
de Facte, un et universel, comme la species ou une chose existant obiective.

Dans la theorie du concept comme fictum , le concept etait ce en quoi se
terminait Facte et ce par quoi l’intellection apprehendait les choses signifiees.
La structure intentionnelle de l’acte intellectif se divisait en deux niveaux,
l’acte et le concept dans lequel l’acte se terminait, du moins pour les concepts
universels. Maintenant, il n’y a plus d’intermediaire entre Facte et ce qui est
apprehende : on peut dire que c’est une structure intentionnelle a un seul
niveau. Ce changement de structure n’affecte pas seulement la composition de
Facte mental, mais ce qui est exige de ses composants. Autrement dit, rien ne
remplace la fonction du fictum , a savoir, etre ce en quoi se termine Facte
d’intellection, c’est cette fonction meme qui disparait. II n’y a pas seulement
un changement dans la theorie concernant la nature du concept, mais aussi
dans la comprehension de la structure de Facte intellectif.

Ce changement explique l’absence, dans la presentation positive de la theo¬
rie du concept comme acte, de l’expression ‘terminate : rien ne correspond a
la fonction d’etre ce en quoi se termine Facte intellectif. La structure inten¬
tionnelle de Facte change. Dans la premiere theorie ockhamienne, ce qui ter¬
minait Facte etait ce par quoi etait apprehende ce dont Facte etait le concept,
du moins pour ce qui est du concept universel 60 ; dans la deuxieme theorie,
rien ne termine Facte intellectif.

Le changement de la structure de Facte mene a une intentionnalite plus
directe, sans intermediaries, ce qui constitue une motivation centrale dans
F evolution ockhamienne. Ce modele d’intentionnalite correspond a un acte
qui est une qualite absolue, et non pas un acte qui ne serait defini que par son
terminus ad quern. Cette caracteristique de la nouvelle conception de la struc¬
ture de Facte intentionnel peut rendre plus acceptable l’idee que sa modification
puisse dependre des considerations purement epistemologiques.

II me semble en effet que la confrontation entre les deux theories doit etre
vue non seulement comme l’opposition entre deux theories concurrentes sur
le statut ontologique du concept, mais aussi comme l’opposition entre deux
theories concurrentes sur la structure meme de Facte, dans lesquelles la fonc¬
tion et les contraintes concernant l’objet de l’intellection ne sont pas les

60) Ord. IV,, I, d. 27, q. 3, p. 242, 15-21.

E. Perini-Santos / Vivarium 45 (2007) 93-112

111

memes. Le choix de la theorie du concept comme fictum ou comme acte est
une partie du choix du cadre theorique plus large qui definit le role, ou les
roles que le concept aura a jouer. Ce changement dans la structure de 1’acte ne
modifie pas l’idee que les concepts seront les composants du langage mental.
Au contraire, comme le remarque Claude Panaccio, cette evolution montre
une intuition profonde d’Ockham que “les divers roles qu’il voulait attribuer
au concept, dans la mesure justement ou ce sont des fonctions semantiques ,
peuvent etre adequatement tenus par 1’acte lui-meme” 61 . Mais les fonctions
semantiques ne sont pas toutes les fonctions qu’un concept peut etre appele a
jouer. Un acte ne termine pas un autre acte intellectif, ni ne se termine
soi-meme, alors qu’un fictum terminait un acte d’intellection qui exigeait pre-
cisement un terminus ad quern.

Une petite question des Quaestiones in Libros Physicorum Aristotelis , dans
une serie consacree a la nature du concept, a pour titre precisement “Le con¬
cept est-il une qualite terminant l’acte intellectif?” 62 . L’argument principal
pour la reponse positive semble n’etre qu’une definition stipulative : le concept
est ce qui est apprehende et termine l’acte intellectif, et ceci doit etre une
qualite de l’esprit 63 . La reponse ockhamienne est negative, et le refus de
l’argument principal est bref, mais instructif:

Sur l’argument principal, je dis que la presupposition est fausse, parce que le concept ne
termine pas 1’acte d’intellection. Car une fois une telle qualite detruite par la puissance
divine, nous aurons toujours un vrai concept des choses. Par consequent, une telle [qualite]
terminant < l’acte intellectif > n’est pas le concept 64 .

La qualite censee terminer l’acte intellectif peut etre detruite sans que le con¬
cept le soit, parce que l’acte peut exister sans cette qualite qui le termine.
Autrement dit, le concept n’est pas le genre d’acte qui ne peut exister qu’avec
quelque chose qui le termine. Un acte qui est termine par quelque chose ne
saurait exister sans quelque chose qui le termine, comme un mouvement ne

61) Claude Panaccio, Les mots, les concepts et les choses , 68.

62) “Utrum conceptus sit qualitas terminans actum intelligendi”, Quaest. Phys., q. 5, pp. 405-406.

63) “Quod sic: Quia conceptus est illud quod concipitur et terminat actum intelligendi ; sed
nihil terminat actum intellectus nisi res extra vel qualitas mentis ; conceptus autem non est res
extra, ut patet ex praedicitis ; ergo etc.”, Quaest. Phys., q. 5, 3-6, p. 405.

M) “Ad argumentum principale dico quod assumptum est falsum, quia conceptus non terminat
actum intelligendi, quia tali qualitate destructa per potentiam Dei, adhuc haberemus verum
conceptum rei. Et per consequens talis [qualitas] terminans non est conceptus.”, Quaest. Phys.,
q. 5, 20-23, p. 406. Cf. Quod., IV, q. 35, 97-133, p. 473.

112

E. Perini-Santos / Vivarium 45 (2007) 93-112

peut exister sans un terminus ad quem. Ces deux phrases ockhamiennes
indiquent le type d’acte qu est le concept, un acte dont la structure n’exige pas
un terme.

devolution dans la theorie ockhamienne du concept ne comprend pas seule-
ment la querelle sur le statut ontologique du concept, elle a trait aussi a la
structure de Facte intentionnel. La difficulte majeure pour voir ce point vient
du fait qu’Ockham n’offre qu’une exposition sommaire de la modification
dans la structure de Facte intentionnel, alors qu’il discute longuement et a plus
d’une reprise le statut ontologique des concepts. II y a deux arguments pour
F acceptation de cette interpretation du changement de la nature de Facte :
(i) un acte ne peut etre termine ni par lui-meme ni par une infinite d’individus,
et plus rien ne semble disponible pour terminer Facte intellectif dans la phi¬
losophic ockhamienne ; done soit la theorie est incomplete, soit il nest
plus necessaire que quelque chose termine Facte ; (n) le vocabulaire associe
a ‘terminate napparait pas dans les expositions de la theorie du concept
comme acte.

Comme j’ai essaye de montrer, ce mouvement de la theorie ockhamienne
peut etre retrace chez d’autres auteurs, depuis Fassociation de Fintellection a
une theorie generate de Facte dans un argument expose par Gauthier Burley,
en passant par la critique de cette association chez Gauthier Burley lui-meme
et chez Gauthier Chatton, jusqu’a la reprise de Fidee, dans une alternative qui
nest plus disponible pour Ockham, chez Pierre d’Ailly. II est clair que Fon
peut estimer peu satisfaisant qu’un changement si important soit passe sous
silence par Ockham lui-meme. On peut toutefois remarquer quau moins un
texte parle d’un concept qui ne se termine en rien 65 . Je crois surtout que c’est
la meilleure faq:on de rendre compte des points (i) et (n) ci-dessus.*

65) Quaest. Phys., q. 5,20-23, p. 406. Notons cependant qu’on ne trouve pas ce genre depression
dans la reprise de l’argument dans Quod., IV, q. 35, 97-133, p. 473.

* ] Une premiere version de ce texte a ete presentee dans un colloque a l’UFRGS a Porto Alegre,
Bresil, en 2004, dont je remercie l’audience pour les commentaires encourageants. Je remercie
aussi le professeur C. H. Kneepkens et la redaction de Vivarium pour les remarques qui ont
permis la correction du texte sur plus d’un point. Je remercie enfin le Conselho Nacional do
Desenvolvimento Cientifico e Tecnologico , Bresil, pour le soutien a la recherche dont ce texte est un
resultat.

BRILL

Vivarium 45 (2007) 113-124

RIUM

www.brill.nl/viv

Francesco Patrizi da Chersos Criticism
of Aristotle’s Logic*

Luc Deitz

Bibliotheque nationale de Luxembourg

Abstract

Francesco Patrizi da Chersos Discussiones peripateticae (1581) are one of the most com¬
prehensive analyses of the whole of Aristotelian philosophy to be published before
Werner Jaegers Aristoteles. The main thrust of the argument in the Discussiones is that
whatever Aristotle had said that was true was not new, and that whatever he had said
that was new was not true. The article shows how Patrizi proves this with respect to the
Organon , and deals with the implications for the history af ancient philosophy in
general implied by his stance.

Keywords

Francesco Patrizi, Aristotle’s Organon , Renaissance philosophy, philosophia perennis

By the middle of the 16th century, the high tide of Platonism that had so
powerfully risen with Ficino one hundred years before, had ebbed down con¬
siderably. Aristotle, often clad in the garb of Averroes, had carried the day,
both in physics and in metaphysics. Committed Platonic philosophers were
few and far between in the late Renaissance, but it is among these that we find
one of the greatest speculative thinkers of all times, and one unduly neglected
by modern historians of philosophy: Francesco Patrizi, called ‘da Cherso’ from
the Croatian island of Cres, where he was born on 25 April 1529. 1

A first draft of the article published below was read as a paper at the joint annual meeting of
the Renaissance Society of America and the Society for Renaissance Studies, United Kingdom,
held at Cambridge (UK), 7-9 April 2005, in a panel devoted to ‘Philosophy in the Renaissance’
organized by Sheila J. Rabin. I would like to record my thanks to Sheila Rabin who kindly
invited me to speak at the panel she had organized, as well as to Lodi Nauta who encouraged me
to write the paper up for publication.

0 Francesco Patrizi da Cherso must not be confused with his homonym, the Italian humanist

DOI: 10.1163/156853407X202539

114

L. Deitz / Vivarium 45 (2007) 113-124

Even by generous Renaissance standards, Patrizi led a colourful life and
managed to be a prolific author at the same time. 2 Being, in turns, a soldier
fighting the Turks, an unsuccessful cotton merchant, a budding physician, a
hapless publisher, a bankrupted trader in manuscripts, and a teacher of Greek
to Tarquinia Molza of Modena, a lady of more than common charms, 3 Patrizi
finally sought, and found, the patronage of the Este family in Ferrara, where
a chair of Platonic philosophy was created at the Studio especially for him.
There he taught from 1578-1591, when Pope Gregory XIV (1590-1591), 4
whom Patrizi had met when they were both students at Padua, invited him
to lecture in Rome. Once again, a chair was established ad personam , and in
1592, Patrizi started teaching Platonic philosophy at the Sapienza under
Gregorys successor, Pope Clement VIII (1592-1605). 5 Unfortunately, Patrizi
soon had to face the opposition of a number of powerful Jesuits (who, as is
well known, held little sympathy for Plato), and his main systematical treatise,
the Nova de universis philosophia (1591), was put on the Index of Prohibited
Books in 1596. 6 Patrizi died in Rome on 6 February 1597, more likely than

Francesco Patrizi of Siena (1413-1494), who wrote a couple of treatises on political philosophy
and held the bishopric of Gaeta.

2) Apart from remarks scattered throughout his works, the main source for our knowledge of
Patrizi’s life is an autobiographical letter, dated 12 January 1387 and addressed to Baccio Valori,
that was first published by A. Solerti, “Lettere autobiografiche di F. Patrizi di Cherso, erudito del
secolo XVI”, Archivio storico per Trieste , TIstria e il Trentino 3 (1886), 275-281; reprinted in:
S. Celia, Francesco Patrizi da Cherso: Pagine scelte (Padua 1965), 37-42, as well as in: V. Premec,
Franciskus Patricijus , Belgrade 1968, [100-104], and in: D. Aguzzi Barbagli (ed.), Francesco
Patrizi da Cherso: Lettere ed opuscoli inediti (Florence 1975), 45-51. Among the more recent
accounts of Patrizi s life, the following deserve mentioning: E. Jacobs, “Francesco Patricio und
seine Sammlung griechischer Handschriften in der Bibliothek des Escorial”, Zentralblatt fur
Bibliothekswesen 25 (1908), 19-47, esp. 20-28; P. Donazzolo, “Franceso Patrizi da Cherso eru¬
dito del secolo decimosesto”, Atti e memorie della Societa istriana di archeologia e storia patria 28
(1912), 1-147, esp. 7-47; and B. Brickman, An Introduction to Francesco Patrizi’s Nova de uni¬
versis philosophia (PhD diss., Columbia University, 1941; Columbia 1941), 10-20.—The most
comprehensive (though not complete for the earlier period) bibliography of works by and on
Patrizi can be found on-line at http://core.ecu.edu/phil/ryane/fphp.htm (site created by Eugene
E. Ryan; last updated on 14 October 2004).

3) The poetess Tarquinia Molza (1542-1617) inspired, inter alia, Patrizi s L’amorosa filosofia (first
ed. byJ.C. Nelson, Florence 1963).

4) I.e., Niccolo Sfondrato (1535-1591).

5) I.e., Ippolito Aldobrandini (1536-1605).

6) On this episode see A. L. Puliafito Bleuel, Francesco Patrizi da Cherso, Nova de universis philo¬
sophia: Materiali per unedizione emendata (Florence 1993), xix-xxvi, with a full survey of earlier
literature.

L. Deitz / Vivarium 45 (2007) 113-124

115

not a desperate man, and he was buried at the church of Sant’Onofrio, close
to where Torquato Tasso had found his ultimate rest two years previously.

Patrizi s thought is both simple and difficult, simple being opposed to com¬
plex, and difficult being opposed to easy. It is simple because Patrizi holds one,
and only one, philosophical tenet—namely the supremacy of Plato and his
predecessors over Aristotle and his followers—, and it is difficult, because, for
modern readers, it is often rather hard, and sometimes just impossible, to fol¬
low his arguments in detail. One of the reasons for Patrizi s difficulty lies in the
fact that he had an unparalleled acquaintance with the entire body of Greek
philosophical literature of all periods, including the most recondite texts, tes-
timonia, or other bits of information found in the most outlandish sources.
None of Patrizis many works could be better suited to illustrate how he
worked and how he argued than his massive Discussiones peripateticae (hence¬
forth: DP) that were published in 1581 during his time at Ferrara. 7

A semantical pointer might not be totally superfluous at this juncture. To a
modern reader, the title word Discussiones is likely to suggest something like a
“debate”, an “enquiry”, or even a “dialogue”. This, however, is not the way in
which Patrizi uses the word, for he quite literally takes it to mean a “striking
asunder” or a “dashing to pieces”—an “analysis”, to use still another transla¬
tion, of Aristotle’s philosophy into its constituent parts with a view to refuting
them one by one. In book 2 of the DP , Patrizi pursues this aim by establishing
the points of agreement between Aristotle and those philosophers that had
preceded him. 8 The main thrust of his argument here is that whatever Aristotle

7) The full title of the DP reads Discussionum Peripateticarum tomi quattuor, quibus Aristotelicae
philosophiae universa historia atque dogmata cum veterum placitis collata, eleganterac erudite decla-
rantur , Basilea: ad Perneam Lecythum, 1581 (repr. with an introduction by Zvonko Pandzic
[Quellen und Beitrage zur kroatischen Kulturgeschichte, vol. 9], Koln [etc.] 1999). The DP
consist of four “tomi”, each made up of several “libri”; throughout this paper, references are
made to “tomus”, “liber”, page and line(s) of the DP. In the quotations as given, spelling and
punctuation have been adapted to modern reading habits.—For a brief summary of the contents
of the DP see L. Deitz, “‘Falsissima est ergo haec de triplici substantia Aristotelis doctrina’. A
Sixteenth-Century Critic of Aristotle—Franceso Patrizi da Cherso on Privation, Form, and Mat¬
ter”, Early Science and Medicine 2 (1997), 227-250, here 228-229.

8) The second “tomus”, on which this paper draws, is inscribed: Aristotelis et veterum philosopho -
rum concordiam continens. Although a number of articles, starting with M. Muccillo, “La vita e
le opere di Aristotele nelle ‘Discussiones peripateticae’ di Francesco Patrizi da Cherso”, Rinasci -
mento 21 (1981), 53-119, have specifically been devoted over the past 25 years or so to Patrizis
exegesis of Aristotle, no detailed account of book 2 has come to my attention.

116

L. Deitz / Vivarium 45 (2007) 113-124

had said that was true was not new, and that whatever he had said that was
new was not true. What is interesting about this is less the foreseeable conclu¬
sion as such, but the way in which it is reached, and, above all the momentous
implications of it as spelled out by Patrizi. It is with these that this paper is
mainly concerned.

The notion that Plato’s and Aristotle’s philosophies were ultimately not
only complementary, but truly concordant, or even identical, was obviously
not new in Patrizi’s days. He starts by pointing out that Porphyry and Simpli¬
cius had held it; that Boethius had famously set out to prove it; and that Pico
della Mirandola, so aptly called the Count of Concord, might have shown this
ultimate agreement between the two philosophers and everybody else, if only
Pope Innocent VIII had allowed him.to carry out his visionary plans and pub¬
licly to defend his 900 theses in i486. 9 But the works of these thinkers were
either no longer extant, or had never been completed, or were nearly unacces-
sible, so that Patrizi, as he puts it, took it upon himself to devote “the morning
hours of 88 days” {DP 2,1,179,42-43) to complete this truly super-human
task 10 left unfinished by those who had preceded him, and to show that Aris¬
totle had ultimately nothing new to say in the fields of dialectics, physics,
metaphysics, ethics and biology. I shall consider here only the proofs adduced
in the field of logic.

9) DP 2,1,179,27-37: “Subiit animum cogitatio, an quid a nobis praeterea praestari posset eo
argumento, quod Porphyrius, vir in philosophia maximi nominis, olim de concordia utriusque
philosophiae, Platonis scilicet et Aristotelis, septem libris condiderat, quodque post ipsum Sim¬
plicius magnus itidem philosophus efficiendum sibi proposuerat, et quod Manilius Boethus [sic]
in animo habuit efficere, ac forte effecit. Quorum tamen virorum labores omnes iniuria tempo-
rum interierunt, quos renovare multis post seculis Io. Picus philosophorum sui temporis phoenix
et promisit, et testimonio Marsilii Ficini confecit: attestatur tamen post obitum ipsius, earn
concordiam repertam novis et obscuris characteribus quibusdam scriptam, qui vix ab ipso Pico
dum viveret legerentur, et a nemine exscribi potuisse.”—The reference to “Manilius Boethus” is
puzzling. Simplicius, in his Commentary on the Categories (on which see n. 12 below), which
Patrizi knew and which he is partly following, repeatedly mentions the Peripatetic philosopher
Boethus of Sidon (who must not to be confused with the Stoic of the same name) as one of his
predecessors. By adding the praenomen Manilius , Patrizi seems to be identifying this Boethus
with Anicius Severinus Manlius [sic] Boethius , who had planned to translate the whole of Plato
and of Aristotle into Latin in order to show the agreement between their respective philosophies
(and, one assumes, their ultimate agreement with the teaching of the Church). By writing
“Boethius” in my translation above, I hope to have done justice to Patrizi s intention, even if I
have obviously betrayed the letter of his text. (On Boethus the Peripatetic see P. Moraux, Der
Aristotelismus bei den Griechen , vol. 1, Berlin 1973, 143-179.)

,0) DP 2,1,179,37-39: “Atque utinam aut horum omnium aut alicuius hac in re scripta extarent:
magno enim nos homunculos hoc labore liberassent.”

L. Deitz / Vivarium 45 (2007) 113-124

117

The Kategoriai are the first in the traditional order of Aristotle’s writings,
and Patrizi starts by giving a detailed account of what he takes to be the philo¬
sophical genealogy of the categories of substance, quantity, quality, and rela¬
tion. 11 In order to pin down their pedigree, he mainly relies on the only
complete scholarly commentary on this work to have come down to us from
antiquity, i.e. that of Simplicius. 12 Quite independently of its intrinsic philo¬
sophic interest, Simplicius’ monumental exegesis is of extraordinary value as a
source for doxographical material, since it contains a great number of quota¬
tions and paraphrases relating to the philosophical controversies that had
arisen around Aristotle’s treatise from the time of Andronicus of Rhodes
onwards—for example the important information that Andronicus had sepa¬
rated the so-called post-predicaments, that is to say chapters lOff of the Cate¬
gories that deal with the different meanings of antikeisthai (to be opposite),
proteron (prior), hama (simultaneous), kinesis (movement), and echein (to
have), from the rest of the treatise. One of the authorities repeatedly referred
to by Simplicius is none other than one Archytas of Tarentum, who suppos¬
edly wrote a treatise in Doric Greek called flepl to> kolQoXov Xoyco, i.e. On
universal predication , or, On categories. And this is where things start to become
interesting.

n) DP 2,1,182, 36-187,52 (here 182,36-41): “Nunc primarias partes, sive dialectices uti Zeno-
nia Socraticaque schola, sive logices, uti Democritus et Aristotele posteriores vocaverunt, in
manus sumamus, ac per earum singula speculemur, si quid Aristoteli consonum cum anteces-
soribus philosophis fuit. Voco autem partes primarias eas, quae de terminis [i.e., the Categories ],
de propositionibus [i.e., the De interpretatione\ , de argumentorum formis [i.e., the Prior Analyt¬
ics^, de methodis [i.e., the Posterior Analytics], de eorundem locis [i.e., the Topics], de elenchis
falsarum argumentationum [i.e., the Sophistici elenchi] tractarit.”

,2) Editio princeps of the Greek text printed by Zacharias Kallierges for Nicolaus Vlastos, Venice
1499 (see BSB-Ink S-407); critical edition by C. Kalbfleisch in Commentaria in Aristotelem
Graeca (CAG), vol. 8 (Berlin 1907); medieval Latin translation by William of Moerbeke (first
revised and edited by Paul of Gene^ano, Venice 1516) edited in two volumes by A. Pattin and
others in: Corpus Latinum commentariorum in Aristotelem Graecorum (CLCAG), vol. 5,1-2 (Lou¬
vain [etc.] 1971-1975); Renaissance Latin translation (based on the editio princeps) by Guillel-
mus Dorotheus edited (as a reprint, preceded by an introduction, of the Venice 1540 edition) by
Rainer Thiel and Charles Lohr in: Commentaria in Aristotelem Graeca. Versiones Latinae temporis
resuscitatarum litterarum (CAGL), vol. 8 (Stuttgart-Bad Cannstatt 1999); English translation in
four volumes by M. Chase (Cat. 1-4), E A. J. de Haas and B. Fleet (Cat. 5-6), B. Fleet (Cat. 7-8),
and R. Gaskin (Cat. 9-15) in: Ancient Commentators on Aristotle (London 2000-2003); a French
translation, directed by I. Hadot, has been under way since 1990 (three volumes published to
date). Only the editio princeps of the Greek text seems to have been known to Patrizi.

118

L. Deitz / Vivarium 45 (2007) 113-124

Modern scholarship—I am here following the editor of the ps.-Archytas,
Thomas A. Szlezak 13 —assigns this treatise fragmentarily preserved by Simpli¬
cius, and miraculously found in its entirety in a 16th century Milanese manu¬
script 14 in 1914, to the 1st century BC, but Patrizi, on the strength of the
Letters 9 and 12 included in the corpus Platonicum , as well as of Diogenes
Laertius’ Life of Archytas, identified this pseudo-Pythagorean philosopher
with no-one else than with the supposed Sicilian correspondent of Plato. 15

So far, so good. However, in order to complicate matters a little bit, it must
be added that Patrizi distinguished this Archytas from another, older one, who
was a direct disciple of Pythagoras. 16 Exactly 20 years before Patrizi published
his Discussiones peripateticae , the Calabrese priest Domenico Pizzimenti had
brought out in Venice a little text similarly ascribed to one Archytas of Taren-
tum, and also dealing with the 10 categories—the nowadays extraordinarily
rare Archytae Tarentini Decern praedicamenta Dominico Pizzimentio Vibonensi
interpreter written in mock-Doric Greek. 17 1 have not seen a copy of this text,
but Patrizi certainly knew it, for he explicitly refers to it under its Greek title
of ’Apxtmn) KaGoAaxoi Adyoi 8exa (DP 2,1,183,7-8).

Patrizi was thus confronted with two startlingly similar things: on the one
hand, with Simplicius numerous quotations from Archytas ITepi tco kolQo'Lov
Any®, and on the other hand with Pizzimenti s edition of a Treatise on catego¬
ries , also ascribed to “Archytas”. He correctly saw, mainly owing to a number
of doctrinal differences, that the two texts could not have been written by the
same author, and he concluded, correctly too, that the Archytas who had com¬
posed the KaGoXixoi taSyoi Sexa was another Archytas than the one so often

13) Pseudo-Archytas, Uber die Kategorien. Texte zur griechischen Aristoteles-Exegese, herausgege-
ben, iibersetzt und kommentiert von Th.A. Szlezak (Berlin 1972). On this text see also
P. Moraux, Der Aristotelismus bei den Griechen, vol. 2 (Berlin 1984), 608-628.

14) Cod. Ambrosianus 23 (A 92 sup.), fol. 123r-136v.

15) The (pseudo-Platonic) Letters 9 and 12 are addressed to Archytas ofTarentum. According to
Diogenes Laertius, Vitae philosophorum, 8,79-80, Letter 12 is an answer to a letter written by
Archytas to Plato, in which the former reports to have been given Ocellus treatise On the Origin
of the Universe (on which see n. 22 below).

16) DP 2,1,182,43-49: ‘Archytae duo fuerunt, ambo Tarentini, ambo quoque de Pythagoreorum
familia. Senior alter, Pythagorae ipsius auditor..., iunior alter..., Platoni contemporaneus.”
Patrizis knowledge of the existence of an “older Archytas” is derived from Iamblichus, Vita
Pythagorae , 104 (Archytas ho presbyteros ).

17) On the printing history of this text see Szlezak, op. cit. (n. 13 above), 7-13. It is edited on

pp. 61-68.

L. Deitz / Vivarium 45 (2007) 113-124

119

quoted by Simplicius. 18 However, working on the assumption that Platos cor¬
respondent was identical with Simplicius’ Archytas, Patrizi incorrectly inferred
(mainly from the treatise’s being written in “Doric” Greek) 19 that the author
of the KotGotaicoi ^oyoi Sera had to be identical with the direct disciple of
Pythagoras, and thus had lived at least 80, if not 120 years before the second
Archytas. 20 Although it is impossible to say when the Ka0oA,iKoi Adyoi Sera
were really written, they were certainly not known to Simplicius, so that the
2nd half of the 6th century AD is a safe terminus post quem for their composi¬
tion, but a much later date seems equally conceivable. 21 In any case, by ascrib¬
ing the text to Archytas, the disciple of Pythagoras mentioned by Iamblichus,
Patrizi dated it at least a generous 1100 years too early.

In order to illustrate what this means in practice, there is no better way than
to quote Patrizi’s own words at the beginning of his exposition of the category
of substance: 22

18) DP 2,1,182,43-49: “Quem libellum non esse Archytae iunioris dare, ex multis lods, a Sim-
plicio ipsius verbis in commentaria translatis... constat: longe a libelli illius et verbis et sententiis
diversus.”

19) DP 2,1,183,8-9: “Dorica lingua uti mos illis fuerat conscriptus, supercilio quodam Pythago-
reo.”

20) DPI, 1,182,51-183,1.

21) Szlezak, op. cit. (n. 13 above), 183, suggests a “Datierung in byzantinische Zeit... als so gut
wie sicher.”

22) DP 2,1,183, 23-36: “Aristoteles de substantia ita scribit: Ad haec primae substantiae, quia
omnibus aliis subjiciuntur, et omnia alia de his praedicantur vel in his sunt, ob id maxime primae
substantiae dicuntur. Locus totus est Archytae, sic enim ipse: Prima quidem ordinata est substantia,
eo quia sola haec subjicitur aliis, et ipsa per seipsam potest cogitari, alia vero non sine hac, vel enim de
ipsa, vel in ipsis subjacentia praedicantur. In eandem sententiam Ocellus habet haec: Et sane prima
ratio, substantia, est perse existens et subsistens, non indigna alterius ad constitutionem in generatione
subjicitur quatenusgenerata est!' —I have failed to understand why Patrizi suddenly, and on several
occasions, refers to the older Archytas as “Ocellus” (e.g., DP 2,1,183,48; 184,2.19.40 et saepp.).
Although there exists a ps.-Pythagorean treatise, also in mock-Doric Greek and supposedly writ¬
ten by another direct disciple of Pythagoras, Ocellus of Lucania (see Iamblichus, Vita Pythagorae ,
267; text edited by Richard Harder, ‘Ocellus Lucanus’ [Berlin 1926; repr. Dublin 1966],
that purports to be the ultimate source for Aristotle’s doctrine of the eternity of the universe,
there is no relationship whatever between the KaGoXucoi Xoyoi 5exa and the De universi natura
of the ps.-Ocellus. To the best of my knowledge, Patrizi (who, according to Harder, op. cit., XXII,
n. 2, knew Ocellus’ treatise in the edition by Count Lodovico Nogarola [ca. 1509-1339], Venice
1559) nowhere explicitly equates the two, but Hellenistic doxography, as illustrated by Censo-
rinus, De die natali , 4,3, and by Lucian, Pro lapsu inter salutandum , 5, seems to have known a
“gemeinsames Urbild”, consisting of the sequence “Ocellus und Archytas und (ihnen folgend)
Aristoteles”, the whole having been set up “in der offensichtlichen Meinung, Archytas folge

120

L. Deitz / Vivarium 45 (2007) 113-124

This is what Aristotle writes about substance [Cat. 5,2 b 15-17]: “First substances are mainly
called first because they underly everything else, and because everything else is predicated
with respect to them or inheres in them.” This whole passage is taken from Archytas [i.e.,
the correspondent of Plato], for he writes as follows [p. 37,16-19 ed. Szlezak]: “Substance
comes first in the sequence of categories, because only substance underlies the other things
and can be thought of as existing per se and independently of everything else, whereas the
other categories cannot be conceived of without substance. Indeed, they are predicated
with respect to substance or as inherent in substances.” Ocellus [= the elder Archytas]
writes, concerning the same topic [p. 61,12-14 ed. Szlezak]: “The first category is sub¬
stance, which is a thing existing and subsisting by itself. It does not need anything else in
the process of generation, for it subsists by itself inasmuch as it is generated.”

When one compares the similarities between the three texts quoted by Patrizi,
the conclusion to be drawn is obvious: on the assumption that the historical
succession was such as Patrizi believed it to have been, one can safely assume
that Aristotle copied his definition of first substance from the younger Archy¬
tas, who was in turn relying on the writings of the older Archytas. Indeed, as
Patrizi puts it in a nutshell: “Eadem fere omnia apud tres hosce celeberrimos
autores” (“nearly everything is the same in the works of these three very famous
authors”). 23

I do not wish to insist on Patrizi s exposition of the categories of quantity,
quality, and relation, 24 for the same argument as the one applied to substance
keeps recurring: far from being the inventor of the ten categories, Aristotle
was relying on the works of his Pythagorean predecessors; indeed, so Patrizi
concludes, “omnis Aristotelicarum categoriarum succus, omnis doctrina...
titulis, dogmatibus, exemplis saepe etiam verbis ipsis... Archytae libris est
conformis et concors” (“the whole pith of Aristotle’s categories and his whole
teaching is in conformity and in accordance with the books of Archytas—with
respect to the titles, the tenets, the examples, and often even with respect to
the very wording”). 25 And although this is nowhere explicitly stated, I think
that, in Patrizi s eyes, only the Pythagorean background could give the correct

seinerseits den Dogmen des Ocellus” (Harder, op. cit., 33 with n. 1). It can only be assumed that
Patrizi was convinced that the elder Archytas and Ocellus were one and the same person without
ever saying so; otherwise I am at a loss to explain this mistaken identification.

23) DP2,1,183, 20-21.

24) Dealt with in DP 2,1,184-187.

25) DP 2,1,186, 41-43. Cp. DP 2,1,187, 30-32: “Itaque dogmata omnia, quae Praedicamento-
rum libro Aristotelis sunt comprehensa, ostendunt maximam Aristoteli fuisse cum antiquis con-
sensionem” (“therefore, all tenets included in Aristotle’s book On categories show that he was in
perfect agreement with the ancients”).

L. Deitz / Vivarium 45 (2007) 113-124

121

answer to the much debated question why Aristotle reckoned with exactly ten
categories, and with neither more nor less: the number 10 corresponds to the
holy tetraktys (1+2+3+4=10) of the Pythagoreans; hence, the ten categories
are nothing but an instance of this all-pervasive structure governing the laws
of visible and of invisible reality.

So much, then, for the Kategoriai. With characteristic thoroughness, Patrizi
next deals with the second of the logical treatises in the traditional order of the
Organon , the Ilepi epjiriveiaq ( Liber de interpretatione) . 26 Right from the start,
Patrizi sides with Andronicus of Rhodes in assuming the Ilepi eppriveiaq to be
a work spuriously attributed to Aristotle, which does, however, not alter the
fact that the reputation that the Liber de interpretatione had gained during the
centuries was linked to the name of Aristotle, and that it must therefore be
subjected to serious scrutiny and criticism. 27

Almost predictably, the first thing to say is that the “ars propositionum” was
already well-known to Plato, who had pointed out in the Sophist that sen¬
tences containing a noun and a verb such as “X is speaking” or “Y is sitting”
are minimal “sermones” or “propositiones”, and that their combination, alter¬
ation or permutation could lead to positive and negative, and in both cases
to true or false utterances. 28 Patrizi similarly points out that Plato had made a
distinction between particular perceptions and universal statements derived
from them in the Theaetetus , so that Aristotle had nothing new to say even in
this respect. 29 Now, particular and universal, true and false, affirmative and
negative, existent and non-existent are all notions of central importance to
anyone remotely interested, as Plato surely was, in clear and consistent thought,
and one might well object to Patrizi that, ultimately, it does not really matter
a great deal whether they were first formulated by Plato or by a first-generation

26) DPI, 1,187,53-188,33.

27) DP 2,1,187,54-56: “Ars vero propositionum, quae libro Periherminias uti Andronicus Rho-
dius et cum eo nos putamus falso Aristoteli tributo continetur, initia ab aliis quam ab Aristotele
coepit.” On the reasons for Andronicus’ rejection of the De interpretatione from the corpus of
genuine Aristotelian writings, and on the ancient authorities for this, see Moraux, op. cit. (n. 9
above), 117-119.

28) DP 2,1,187,45-188,8: “Nam Plato in Sophista propositionem Xoyov vocans, uti postea Topi-
corum author, ait ‘Homo discit’ [Sophist. 262c], ‘Theaetetus sedet’ [Sophist. 263a] esse sermonem
minimum docetque ex nomine et verbo componi... Novit etiam longiores propositiones...
Novit quoque ‘sermonem alium esse verum, alium falsum’ [Sophist. 263d and 264a]... Affir-
mationem quoque et negationem..—‘Noun’ and ‘verb’, ‘true’ and ‘false’, ‘affirmation’ and
‘negation’ are the key concepts introduced in the opening paragraph of the De interpretatione.

29) DP 2,1,188,10-11: “Universalia quoque et particularia alia esse in Theaeteto docet.”

122

L. Deitz / Vivarium 45 (2007) 113-124

reader of his works. Such an objection, however, could only be made by
a contemporary reader with an extraordinarily limited outlook on the true
history of ancient philosophy, that is to say by a reader ignorant of the fact,
well-known to Patrizi, that, long before Plato, all the logical and metaphysical
niceties just hinted at had already been set down by another Pythagorean, a
lady by name of Periktione to whom are attributed two works fragmentarily
preserved by Stobaeus: one On Wisdom , written in pseudo-Doric, and the
other On the Adornment of Women, written in pseudo-ionic. On the strength
of one single sentence preserved in the fragment On Wisdom , Periktione—
who, incidentally, bears the same name as Plato’s mother, who was herself
supposed to have been a Pythagorean—is thought by Patrizi to have been
the real inventor of the kind of propositional logic contained in the Ilepl
eppriveiaq. Here is what Patrizi takes to be proof of this: 30 “Accidents of things
are either universal and common to all, or common to most of them, or they
belong only to one single thing.”—“I know full well”, Patrizi adds, “that these
words do not cover the whole ars propositionum’ contained in the Ilepl
eppriveiaq, but they are the elements of the whole of this art, and they show to
what extent Aristotle wished to be in agreement with the ancients on this
point too.” 31

It would be equally pointless to argue on historical grounds that the frag¬
ments ascribed to Periktione were certainly written after Aristotle’s time, or
on systematical grounds that—even on the assumption that Aristotle had
indeed been able to avail himself of Periktione’s On Wisdom —the one sen¬
tence quoted by Patrizi can hardly suffice to prove Aristotle’s acquaintance
with, or dependence on, this spurious work in his De interpretatione. Indeed,
any such objection would not do justice to Patrizi’s primary intention in the
opening chapter of book 2 of the Discussiones peripateticae , which is to show

30) DP 2,1,188,19-22: “Perictione Pythagorea libro De sapientia: ‘Accidentia entium quaedam
universaliter accidunt omnibus, quaedam plurimis ipsorum, quaedam etiam uni singulari’.”
Nothing is known about this Periktione, who is neither mentioned by Diogenes Laertius nor by
Iamblichus in his Vita Pythagorae, and whose name occurs only three times in the Anthologium
of Stobaeus. For the fragment referred to by Patrizi see I. Stobaeus, Anthologium , rec. C. Wachs-
muth et O. Hense, 5 vols. (Berlin 1885-1912; 2nd repr. Hildesheim [etc.] 1999), 3,1,121
(= vol. 3, p. 86). The text can also be found in F. W. A. Mullach, Fragmenta philosophorum Grae¬
corum, 3 vols. (Paris 1860-1881; repr. Aalen 1968), vol. 2, 32-33, and in H. TheslefF, The Pythag¬
orean Texts of the Hellenistic Period (Abo 1965), 146.

30 DP 2,1,188,25-28: “Scio mehercule haec non totam propositionum artem, quae libro Peri-
herminias continetur, amplecti, attamen principia eius totius artis sunt ostenduntque quantum
Aristoteles (si modo suus is liber est) in hac quoque re cum antiquis concors esse voluerit.”

L. Deitz / Vivarium 45 (2007) 113-124

123

how Aristotle was dependent on, or at least sought to be in agreement with,
those philosophers who had preceded him.

In view of what has been said so far, I think it is enough just to mention the
three “sources” singled out by Patrizi for the remaining works of the Organon ,
without entering into technical details. Indeed, in all cases, the “proofs”
adduced are very similar to the ones that we have become acquainted with by
now. Thus, for the third and the fourth works in the traditional order of Aris¬
totle’s logical writings, the Prior and the Posterior Analytics that contain Aristo¬
tle’s syllogistics, Patrizi is able to pin down yet another ps.-Pythagorean text as
their ancestor, the Libellus de mundo of the ps.-Timaeus Locrus. 32 With respect
to the Topics , he maintains that one of its key notions, i.e. that of “definition”,
was certainly not inventend by Aristotle, nor by Plato, but once again by
Pythagoras himself. 33 And the mode of reasoning set out in the Sophistici den-
chi, finally, goes back at least to Zeno of Elea. 34 Thus, in Patrizi’s eyes, there is
only one conclusion that can be drawn from his analysis of Aristotle’s Orga¬
non : 35 “Be this as it may, I think that it is perfectly evident that the beginning
and many developments of logic [...] had been written by others before Aris¬
totle, and that a great deal of agreement with Aristotle can be found in their
writings [...].”

However much mistaken Patrizi’s premises may have been when they are
judged according to the standards of a modern reader, the only possibility for
this selfsame reader to do Patrizi full justice is to measure him according to his
own standards by taking him seriously in what he says, and, as a philosopher,
to judge him according to the honesty, consistency, and completeness of his
inferences. Since there can be no reasonable doubt concerning the latter, we
can only conclude that, by repeatedly placing Aristotle in the lineage of Pytha¬
goras and his successors, Patrizi followed a treble aim:

32) DP 2,1,189,49-52: “Timaeus quin etiam Locrus Pythagoreus libello De mundo... Quasi
idem illi sit... quod Aristoteli... in syllogismi definitione.”—Critical edition of the ps.-Timaeus
Locrus by W. Marg, De natura mundi et animae (Leiden 1972).

33) DP 2,1,181 [recte 191], 44-45: “Quid vero de methodis quattuor dicendum? Est sane apud
Laertium in Pythagora ex Phavorini testimonio Pythagoram primum omnium definitione
usum.”—Patrizi s authority here is Diogenes Laertius, Vitaephilosophorum , 8,48.

34) DP 2,1,196,16-17: “Zeno Eleates, qui... author fuerit eristices.” Patrizi quotes Aetius; for
this text from Aetius see H. Diels, Doxographi Graeci (Berlin 1879; repr. 1965), 601.

35) DP 2,1,196,22-26: “Quomodocumque id vero sit, certe et initia et multos progressus et
logices et dialectices utriusque et eristices et categoriarum et propositionum et argumentorum et
methodorum ante Aristotelem fuisse ab aliis conscriptos, in iisque multam cum antiquioribus
Platoneque Aristotelis concordiam clarissime, ut opinor, constitit fuisse.”

124

L. Deitz / Vivarium 45 (2007) 113-124

First, as I have already hinted at, in order to show that Aristotle was not that
solitary summit of philosophy that his admirers took him to be, it was neces¬
sary to prove that all he had to say had already been said, and often said much
better, by others before him. In fairness to Patrizi, one ought to specify that he
did not do this primarily with a view to denigrating Aristotle’s merits, but in
order to show that the insights that he had supposedly had, or the truths that
he had supposedly discovered, were but a rehash of something much older, an
instantiation, as it were, of the one and indivisible “philosophia perennis” of
whose existence Renaissance Platonists from Ficino onwards were so strongly
convinced.

Second, by directly linking Aristotle s Organon to the pure spring of Pythag¬
orean wisdom, Patrizi is able to derive its ultimate, i.e. pre-Pythagorean, origin
from the only true well- and fountainhead of all wisdom and philosophy, who
is none other than Hermes Termaximus. Lest one should think that little logic
in the strict sense is to be found in the Corpus Hermeticum , let us remem¬
ber—and Patrizi himself clearly points this out—that the second treatise
of the Hermetica , the one following upon the Poimandres , bears the title
KaBoXucoq Adyoq. 36 This is so patently reminiscent of the titles with which the
two Axchytases inscribed their own Libelli , 37 that one cannot possibly speak of
a coincidence, but has to assume that at least the elder Archytas derived the
designation he used for the categories— kcx 06 A,od Xoyoq —directly from the
treatise of Hermes, who must thus be reckoned to have been the true and
ultimate inventor of dialectics.

Third and last, by finally restoring what he took to be the true' genealogy
of one extraordinarily important segment of ancient philosophy, Patrizi could
not but see himself as one of those who contributed to the revival of pristine
learning, and hence as one of those who contributed to the advancement of
mankind in general. Patrizi clearly was a man with a mission, and by proving
that Aristotle was dependent on Pythagoras, and thus ultimately on Thrice
Greatest Hermes, he took one important step towards fulfilling this delicate
mission.

36) DP 2,1,188,24-25: “Hermetis Termaximi extat libellus, qui Kcc0oA,ik6<; Tvoyoq titulum prae-
fert.”

37) See notes 13 and 17 above.

Vivarium 45 (2007) 125-127

BRILL

VIVA

RIUM

www.brill.nl/viv

Review

Paul Vignaux, Philosophie au Moyen Age. Edite, presente et annote par R. Imbach. Vrin: Paris
2004. ISBN 2-7116-1680-0.

The main text is a reprint of a book from 1987, which was a reprint of one from 1958, which
was a revised version of one from 1938. The 1938 edition significantly did not mention philoso¬
phy in its title, it was called La pensee au Moyen dge, and I think this was a less misleading title
than the later one.

The 2004 edition starts with an interesting sketch by Imbach of Vignaux in context both as a
historian of medieval philosophy and as a trade unionist. Next follow Vignauxs own long intro¬
duction to the 1987 edition and his shorter foreword to the 1958 edition. The title of chapter I
‘Renaissance, humanisme’ refers to the Carolingian renaissance and that of the 12th century.
Alcuin, Eriugena, Thierry of Chartres and other well-known persons appear there. Chapter II is
‘Quatre fondateurs: Saint Anselme, Abelard, Saint Bernard, Richard de Saint-Victor. Chapter II
‘Conditions du XIII e siecle’, IV ‘Diversite du XIII e siecle: Robert Grosseteste et Roger Bacon,
Saint Bonaventure, Saint Thomas et son temps, Hors du thomisme’ (under the last item we find
Lullus, among others), V ‘Jean Duns Scot, Guillaume d’Occam’, VI ‘Aspects des XIV e et XV C
siecles’. Finally, the volume includes the hitherto unpublished essay ‘Histoire de la pensee
medievale et problemes theologiques contemporains’ that Vignaux wrote in 1987.

Both the 1987 introduction and the concluding essay from the same year, the year in which
Vignaux died, try to explain the author’s approach to medieval thought. Born in 1904, his for¬
mative years were the period between the two world wars. A catholic and a philosopher, a histo¬
rian and a man with an active engagement in the social and political issues of his time, he tried
to find a way to be all of these things in a coherent way. The change of title from La pensee to
Philosophie reflects, he tells us (p. 289), his final conviction that there is such a thing as a phi¬
losophy of religion, in which one applies rationally articulated concepts to a historically given,
positive religion. This is what he sees medieval thinkers doing, though he is careful not to make
them all variants of the same general type, time and again stressing the diversity of medieval
thought. As he very nicely puts it, “on ne ramene pas une epoque a une essence” (p. 103). With
each author he tries to to identify “l’ordre des intentions, qui donnent l’unite aux pensees
vivantes” (p. 186) rather than a timeless system.

It is not difficult to understand why Vignauxs stress on the diversity could feel refreshingly
new to a catholic public raised on the cult of “Thomas, Thomas fiber Alles”, and the book cer¬
tainly does contain many good and astute observations and interpretations. Still, I must admit
that it is incomprehensible to me why anyone would think a reprint was needed in the beginning
of the twenty-first century.

Vignaux corrected many of the excesses of Neo-scholasticism, and his Catholicism was not
narrow-minded: he was interested in Luther and sympathetic to Karl Barth, yet he stayed within
a fundamentally religious universe determined by his catholic background and the problems

DOI: 10.1163/156853407X183199

126

Review / Vivarium 45 (2007) 125-127

facing intellectual catholics of his age. The interior Christ and the triune God are (is?) all over
the book, and Vignauxs prose can grow quite passionate when dealing with the subject; for
someone with a secular approach to the Middle Ages several key passages are barely intelligible.
In spite of its opposition to a then fashionable sort of medieval philosophy written by believers
for believers, this book is itself one written by a believer for co-believers. Indeed, Vignauxs
universe of thought is in some ways less comprehensible to me than that of many medieval
thinkers.

The book was written at a time when the shadow of Etienne Gilson loomed large, and to a
considerable extent it is a debate with Gilson—but who now cares about Gilson and thinks a
debate about his views is needed? Can the book still feel relevant at least to a young French
catholic with an interest in medieval thought? Being neither young nor French nor a catholic it
is difficult for me to judge, but I should think that most young French catholics would feel that
Vignaux addresses problems that have never been vital to them, though they would have been
vital had they lived a generation or two earlier.

As opposed to many of his contemporaries, Vignaux was well aware of the centrality of gram¬
mar and logic to medieval philosophers and theologians, saying, e.g., “La technique logico-
grammaticale constitue un trait durable de Thomme medieval” (p. 105). Yet, logic and grammar
hardly play any role in his book, and he seems to have had some difficulty in understanding why
anyone could take a serious interest in such childish matters, speaking (p. 104) of “les jeux form-
els [of logic and grammar], si loin de notre mentalite et pour nous sans interet” and on p. 109,
in a passage about Abelard and the twelfth century, we read “Une logique qui a perdu pour nous
toute fraicheur enchanta des hommes qui la decouvraient”. On p. 231 it is said that Ockham
formulated the problem of universal “a partir de la theorie de la supposition, and then we are
exhorted to try to play along with Ockham: “Mettons-nous par un example dans l’atmosphere
logico-grammaticale oil vivait et se plaisait son esprit”; we are obviously not supposed to feel that
serious philosophy can thrive in such an atmosphere, nor are we much encouraged to change our
mind. At the time when Vignaux wrote those words most potential readers would happily
include themselves among his nous. A Philotheus Boehner might think otherwise, but there were
not many like him in those days, and “formalism” was frowned upon by traditional Neo¬
scholastics (presumably because it was characteristic of such godless people as Betrand Russell
and the logical positivists). In 2004 many students of medieval philosophy and theology would
decidedly want to distance themselves from inclusion in Vignauxs nous. Not only do we recog¬
nize that the medievals spent much time on logic and linguistic thought, we also think their
work in the area was extremely interesting, and that it is absolutely necessary for us to under¬
stand it if we want to understand what they say about other philosophical matters.

The lack of serious attention to logic and linguistics is just one of the several reasons why the
book must be considered to be outdated. Many of the issues that were contentious at the time
of writing have ceased to be so—partly through victory to Vignauxs side, partly because few
people care anymore. Further, for obvious reasons the book takes no notice of the many and
sometimes very important text-editions and scholarly studies that have appeared during the
last fifty years. Some factual information is simply false or at least misleading—thus on p. 109
the Ars Nova is described as consisting of “Analytiques, Topiques, Sophismes ”, all translated
by Boethius, while, in fact, of the two Analytics only the Prior Analytics is by him, the Posterior
Analytics being a gift from James of Venice—I have not checked, but wonder whether this is a
left-over from the original 1938 edition of the book; by 1958 Vignaux could have known better,
had he read the relevant articles by Lorenzo Minio-Paluello. The name Sophismes instead of

Review / Vivarium 45 (2007) 125-127

127

Refutations sophistiques once more suggests that logic was not Vignauxs strong side. Nothing is
said about James of Venice’s translations of non-logical works, in fact we are told that De anima
and the other works of natural philosophy only came later via the Arabs (see also p. 147).

Imbach’s notes and bibliography are insufficient to update the old book. The notes do not, for
instance, rectify the erroneous information about the Aristoteles Latinus , nor do they modify the
characterization (p. 145) of grammatica speculativa and modi significandi as “Ces matieres logico-
grammaticales, a peine explorees” by referring to the abundant bibliography on the subject from
the 1960s and onwards. (To Vignauxs credit, it should be noted that he had understood that the
modi significandi were important, he just knew too little about them.) In Imbachs bibliography
one misses some of the most important books that Vignaux could not know or had neglected.
Thus under Logique, semantique de l'argumentation , theorie de la connaisance we find neither Jan
Pinborg s Logik und Semantik im Mittelalter nor Alfonso Maieru s Terminologia logica della tarda
scolastica , both from 1972.

I am sad to have had to be so negative in my evaluation of what was once a very good book.
But in my view, it is hopelessly outdated and only interesting for a historian of the history of
medieval thought or for a historian of the history of twentieth-century catholic thought.

Copenhagen Sten Ebbesen

Vivarium 45 (2007) 128-130

BRILL

VIVA

RIUM

www.brill.nl/viv

Books Received

George Arabatzis, ‘Paideia and ‘Episteme in Michael of Ephesus. In De part. anim. I, 1,3-2, 10
(Athens: Academy of Athens, Research Center on Greek Philosophy, 2006). 340 pp. ISBN
960-404-092-8. [in Greek, with English summary].

Adriano Oliva, Les Debuts de Tenseignement de Thomas d’Aquin et sa conception de la ‘Sacra
Doctrinal, avec Sedition du prologue de son commentaire des Sentences (Paris: Vrin, 2006).
416 pp. ISBN 2-7116-1827-7.

Nicolas dAutrecourt et la faculte des arts de Paris (1317-1340). Actes du colloque de Paris 19-21 mai
2005, eds. S. Caroti et C. Grellard (Cesena: Stilgraf Editrice, 2006).

Contents: William J. Courtenay, “Arts and Theology at Paris, 1326-1340”; Dallas G.
Denery II, “Nicholas of Autrecourt on Saving the Appearances”; Dominik Perler, “Rela¬
tions necessaries ou contingentes? Nicolas d’Autrecourt et la controverse sur la nature des
relations cognitives”; Aurelien Robert, “Jamais Aristote n’a eu de connaissance d’une sub¬
stance: Nicolas d’Autrecourt en contexte”; Christophe Grellard, “Nicolas Drukken de
Dacie, entre Autrecourt et Buridan. Logique et theorie de la connaissance a la fin des
annees 1330”; Jack Zupko, “Buridan and Autrecourt: A Reappraisal”; Jean Celeyrette,
“L’indivisibilisme de Nicolas d’Autrecourt dans la contexte parisien des annees 1330”; Ste-
fano Caroti/“Nicolas d’Autrecourt, la generation, la corruption et l’alteration”; Jean-
Baptiste Brenet, “Averroes et les ‘averroi'stes’ dans le traite ‘Sur l’eternite des choses’ de
Nicolas d’Autrecourt”; Guido Alliney, “The Theory of the Will in Nicholas of Autrecourt:
A Threefold Structure”; Joel Biard, “Nicholas d’Autrecourt et Gautier Burley”.

Theological Quodlibeta in the Middle Ages. The Thirteenth Century, ed. Christopher Schabel
(Leiden: Brill, 2006). 363 pp. (Brill’s Companions to the Christian Tradition, 1). ISBN-
13: 978-90-04-12333-5; ISBN-10: 90-04-12333-4.

Contents: J. Hamesse, “Theological Quaestiones Quodlibetales' ; Kevin White, “The Quod¬
libeta of Thomas Aquinas in the Context of His Work”; Girard J. Etzkorn, “Franciscan
Quodlibeta 1270-1285”; Hans Kraml, “The Quodlibeta of William de la Mare”; Pasquale
Porro, “Doing Theology (and Philosophy) in the First Person: Henry of Ghent’s Quodli-
beta”\ Giorgio Pini, “Giles of Rome”; John F. Wippel, “Godfrey of Fontaines’ Quodlibet
XIV on Justice as a General Virtue”; Elsa Marmursztejn, “A Normative Power in the Mak¬
ing”; Sylvain Piron, “Franciscan Quodlibeta in Southern Studia and at Paris, 1280-1300”;
Roberto Lambertini, “Political Quodlibeta ’; Giovanni Ceccarelli, “‘Whatever’ Economics:
Economic Thought in Quodlibeta ’; Jean-Luc Solere, “Was the Eye in the Tomb?”.

Nadja Germann, De Tempore ratione. Quadrivium und Gotteserkenntnis am Beispiel Abbos von
Fleury und Hermanns von Reichenau (Leiden: Brill, 2006). 384 pp. 15 plus plates. (Studien
und Texte zur Geistesgeschichte des Mittelalters, 89). ISBN-13: 978-90-04-15395-0;
ISBN-10: 90-04-15395-0.

DOI: 10.1163/156853407X202548

Books Received / Vivarium 45 (2007) 128-130

129

Guillelmi de Luxi Postilla super Baruch. Postilla super Ionam , cura et studio Andrew T. Sulavik
(Tumhout: Brepols, 2006). xcii +178 pp. (Corpus Christianorum Continuatio Mediaeva-
lis 219). ISBN 978-2-503-05191-8.

Ecriture et reecriture des textes philosophiques medievaux. Volume dhommage offert a Colette Sirat ,
ed. J. Hamesse et O. Weijers (Turnhout: Brepols, 2006). 499 pp. (Federation Internatio¬
nale des Instituts d’Etudes Medievales, Textes et Etudes du Moyen Age, 34). ISBN 978-2-
503-52424-5.

Contents: List of C. Sirat’s publications; 23 contributions by: A. Chahlane; G. Dahan;
S. Di Donato; I. Dobbs-Weinstein; N. Elsakhawi; G. Federici Vescovini; M. Geoffrey;
R. Glasner; J. Hamesse; S. Harvey and Ch. H. Manekin; W. Zeev Harvey; R. Hissette;
A. L. Ivry; S. Klein-Braslavy; H. Kreisel; D. J. Lasker; J. Marenbon; M. C. Monteiro Pacheco;
D. Poirel; J.-P. Rothschild; R. Schmidt van Gelder-Fontaine; O. Weijers; M. Zonta.

Intellect et Imagination dans la philosophie medievale = Intellect and imagination in medieval phi¬
losophy = Intelecto e imaginagdo na filosofia medieval: Actes du Xle Congres international de
philosophie medievale de la Societe Internationale pour lEtude de la Philosophie Medievale
(S.I.E.P.M.), Porto du 26 au 31 aout 2002 , ed. Maria Candida Pacheco et Jose Francisco
Meirinhos (Turnhout: Brepols, 2006), 3 vols., 2008 pp. (Societe Internationale pour
l’fitude de la Philosophie Medievale. Rencontres de Philosophie Medievale, 11). ISBN 2-
503-51818-4 (3 vol. set).

Dan Burton, Nicole’s Oresmes \De visione stellarum (On Seeing the Stars). A Critical Edition of
Oresmes Treatise on Optics and Atmospheric Refraction , with an Introduction , Commentary
and English Translation (Leiden: Brill 2007). 319 pp. (Medieval and Early Modern
Science, 7). ISBN 90-04-15370-5.

Cahiers de llnstitut du Moyen-Agegrec et latin, 77 (2006), 155 pp. ISSN 0591-0358.

Contents: Istvan Hajdu, “Vita sancti Ladislai confessoris regis Hungariae”; Magdalena
Bieniak, “A Critical Edition of Stephen Langton’s Question De persona "; Ernesto Santos,
“Is 'Deus scit quicquidscivit an Epistemic Sophisma?”; David Bloch, “Averroes Latinus on
Memory. An Aristotelian Approach”; David Bloch, “The Aldine Edition of Aristotle’s De
sensu .

Documenti e studi sulla tradizione filosofica medievale , XVII (2006) 599 pp. ISSN 1122-5750.
Contents: Anthony J. Celano, “The Understanding of Beatitude, the Perfection of the Soul
in the Early Latin Commentaries on Aristotle’s Nicomachean Ethics ”; Martin J. Tracey, “An
Early 13th-Century Commentary on Aristotle’s Nicomachean Ethics ” 1,4-10: The Lectio
cum Questionibus of an Arts-Master at Paris in MS Napoli, Biblioteca Nazionale, VIII G 8,
ff 4r-9v”; Tobias Hoffmann, “Voluntariness, Choice, and Will in the Ethics Commentaries
of Albert the Great and Thomas Aquinas”; Silvana Vecchio, “II discorso sulle passioni
nei commenti all ’Etica Nicomachea : da Alberto Magno a Tommaso d’Aquino”; Martin
W. F. Stone, “Equity and Moderation: The Reception and Uses of Aristotle’s Doctrine of
< £7ti£iK£i(x in the Thirteenth Century Ethics”; Iacopo Costa, “II problema dell’omonimia
del bene in alcuni commenti scolastici dNEtica Nicomachea "; Luca Bianchi, “Boece de
Dacie et PEthique a Nicomaque"\ Sonia Gentili, “L \ Etica volgarizzata da Taddeo Alderotti
(m. 1295). Saggio di commento”; Virpi Makinen, “The Influence of the Commentaries on
Aristotle’s Nicomachean Ethics and Politics on the Discussion on Property Rights”; Paolo
Falzone, “Ignoranza, desiderio, giudizio. L’ Etica Nicomachea nella struttura argomentativa
di Monarchia, III 3”; Gianfranco Fioravanti, “Etica e biologia in un anonimo trattato di
eugenetica. Edizione del Libellus de ingenio bone nativitatis (ca. 1314)”; Luciano Cova,

130

Books Received l Vivarium 45 (2007) 128-130

“Felicita e beatitudine nella Sententia Libri Ethicorum di Guido Vernani da Rimini”; Taina
Holopainen, “The Will and Akratic Action in William Ockham and John Duns Scotus”;
David A. Lines, “Pagan and Christian Ethics: Girolamo Savonarola and Ludovico Valenza
on Moral Philosophy”; Sigrid Muller, “Wiener Ethikkommentare des 15. Jahrhunderts”;
Guido Alliney, “Per un con fronto fra le redazioni del Commento alle Sentenze di Francesco
delta Marchia: la versione ‘maggiore’ di In Sent., I, d. 1, q. 6”; O. Nielsen, Cecilia Trifogli,
“Questions on the Beatific Vision by Thomas Wylton and Sibert de Beka”; Indice dei
Manoscritti; Indice dei Nomi.

Sacris Erudiri. A Journal on the Inheritance of Early and Medieval Christianity, 45 (2006),
489 pp. ISSN 0771-7776.

Contents: “In memoriam Fernand Bossier”; C. T. Begg, “The Youth of Samuel”; B. Selter,
“Astral immortality in the Carmina Latina Epigraphica of the City of Rome: A Comparison
between Pagan and Christian Views”; M.-P. Bussieres, “L’influence du synode tenu a Rome
en 382 sur l’exegese de YAmbrosiaster” ; R. Gonzalez Salinero, “Retorica y violencia contra
los judios en el Imperio cristiano (siglos iv y v)”; G. Folliet, “La double tradition patristique
du verset Job 28, 28: « Pietasltimor Domini est sapientia. .. »”; F. Dolbeau, “Brouillons
et textes inachev^s parmi les oeuvres d’Augustin”; J. P. Crowley, “Latin Prayers Added into
the Margins of the Prayerbook British Library, Royal 2.A.XX at the Beginnings of the
Monastic Reform in Worcester”; R. Kottje, “Bussbiicher in mittelalterlichen Bucherver-
zeichnissen”; C. Wollin, “Beitrage Zur Werkchronologie und Rezeption des Matthaus von
Vendome”; W. Ysebaert, “Cinq lettres inconnues d’fitienne de Tournai (1128-1203)”;
J. Donnadieu, “ CHistoria orientals de Jacques de Vitry. Tradition manuscrite et histoire du
texte”; Th. Haye, “Der Krieg um Troja als Kampf der literarischen Methoden. Eine histo-
rische und poetologische Analyse des Gerhard von Liittich aus dem Jahr 1373”; Opera ad
redactionem transmissa; Index codicum.

BRILL

Vivarium 45 (2007) 131-135

VIVA

RIUM

www.brill.nl/viv

Introduction

A common way of dividing up medieval logic, which goes back to the period,
is into the logic vetus , the logic a nova and the logica modernorum. The first two
of these labels refer to the different items in Aristotle’s logical corpus—the
group of works that came into use gradually between c. 780 and c. 1000, and
the remaining works that became known in the century or so after 1110. The
third is the title for those branches of logic invented by medieval thinkers,
such as the theory of the properties of terms or the investigation of insolubles.
This division sums up very eloquently a conception of medieval logic that is
usually unquestioned. Either it is Aristotelian in its basis, or else it is—perhaps
with some debts to the Aristotelian tradition—of medieval invention. But
might there not be other ancient traditions behind aspects of medieval logic?
Might it not have many ancient roots? The papers here set out to explore some
of the avenues indicated by this question. TBe sections of the collection indi¬
cate the ways in which the question can be articulated.

The other great tradition of logic and semantics in the ancient world besides
Aristotle’s was that of the Stoics. There is no doubting its importance as a
logical school, but, unlike Aristotelian logic, its transmission to the Middle
Ages is unclear. Section II (papers on conditionals by Christopher Martin; on
compositionality by Martin Lenz and on res meaning a thing thought by Anne
Grondeux) considers the extent of possible Stoic influence. By contrast, there
is no doubt that medieval thought was influenced by Platonism, but there is a
large question—addressed in Section III (papers on Eriugena by Christophe
Erismann, Ideas in the mind of God by Irene Rosier-Catach and symbolism
in the early Middle Ages by Stefania Bonfiglioli and Costantino Marmo)—as
to whether it is helpful to speak of Platonic logic. One of the main bearers
of Platonic influence was Augustine, but, especially in semantics, Augustine
is strongly influenced by Stoicism. How did this Augustinian approach to
semantics interact with the Aristotelian model, known especially through his
On Interpretation ? The papers in Section IV (Mary Sirridge and Claude Panac-
cio on mental language, and Laurent Cesalli on the semantics of propositions)
examine this question.

Koninklijke Brill NV, Leiden, 2007

DOI: 10.1163/156853407X217678

132

Introduction / Vivarium 45 (2007) 131-135

From Sections I-IV, it will already have become evident that the Aristotelian
ideas combine in a whole variety of ways with these other Stoic, Platonic and
Augustinian ones. Section V looks at what might be called Aristotelian tradi¬
tions in medieval logic—medieval developments in logic that can indeed be
traced back to Aristotle, but to aspects of Aristotle from which theories that are
themselves unAristotelian are developed: the theory of transcendental terms,
discussed by Luisa Valente; metaphor, treated by Jennifer Ashworth and Chris-
tophe Grellard’s examination of how sceptical arguments found in the Prior
and Posterior Analytics are exploited in the fourteenth century. The final two
pieces in this section—by Catarina Dutilh Novaes on fallacies and supposition,
and Egbert Bos on probationes terminorum —consider to what extent branches
of the logic a modernorum can be thought of as distant types of Aristotelianism.

* * *

Sten Ebbesen sets the scene, in Section I, by presenting his conception of
the problem as a whole: ‘The Tradition of Ancient Logic-cum-Grammar in the
Middle Ages—What’s the Problem?’, he asks in his title. Ebbesen is sceptical
about the direct influence on medieval logic of ancient Stoicism. Where Stoic
and medieval logical notions (for instance, Stoic lekta and medieval dicta or
enuntiabilia) are close to one another, he believes that it is a case of parallel
development. He also doubts whether there is anything that can properly be
called ‘Neoplatonic Logic’. But he suggests that we add to the traditional ter¬
minology of ‘Aristotelian’, ‘Stoic’ and ‘Neoplatonic’, the ideas of ‘Hellenistic
Standard’ and ‘Late Ancient Standard’. Hellenistic Standard is the mixture of
Stoic and Aristotelian logic that was widely disseminated in the Hellenistic
period. Ebbesen believes that it would be worthwhile to try to describe it, but
that it has ‘scant relevance’ for medieval logic. By contrast, he argues that Late
Ancient Standard logic and Late Ancient Standard grammar are both impor¬
tant for the Middle Ages: LAS logic is the mixture of logical ideas, often devel¬
oped by Neoplatonists though not specifically Neoplatonic, that was common
in late antiquity and is reflected in Boethius’s works. LAS grammar is largely
but not entirely Stoic in its derivation, and it is mediated through Apollonius
Dyscolus and Priscian.

Ebbesen’s answer to the question he poses himself faces, therefore, in
two different directions. On the one hand, he does not reject the idea that
there are traditions at the basis of medieval logic other than the Aristote¬
lian. Indeed, his argument points in the direction of saying that we should
think, first and foremost, of LAS logic and grammar—composite traditions,

Introduction / Vivarium 45 (2007) 131-135

133

containing heterogeneous elements—as giving medieval logic and semantic its
roots, as multiple as these traditions themselves. This analysis fits a number of
the papers that follow. Grondeux, Lenz and Ashworth illustrate the pervasive¬
ness of LAS grammar, as do the papers in Section IV (where Augustine’s
importance as a transmitter of strands in this tradition is highlighted); Valente,
especially, provides an instance of LAS logic and how it is developed in the
Middle Ages. On the other hand, Ebbesen wants to disappoint some of the
expectations raised by the title of this collection. We should not expect, he tells
us, to encounter medieval versions of Stoic logic, or that of any other ancient
school apart from the Peripatetics. Is Ebbesen right?

In several important ways, the papers published here seem to show that he
is wrong. First, Chris Martin’s paper shows that it is possible, if in a qualified
way, to talk of a medieval influence of strictly Stoic logic. Stoic logic is most
sharply distinguished from Aristotle’s by being propositional, whereas Aristo¬
telian logic is of terms. Unfortunately, ancient scholars seem to have seen these
systems as competing, rather than complementary, and in late antiquity Stoic
logic all but disappeared, leaving no texts that could be transmitted to the
Middle Ages. In LAS logic, as represented by Boethius, although features orig¬
inating from Stoic propositional reasoning remain, their meaning has been
lost (cf. Speca, 2001). What Martin shows, however, is that Abelard succeeded
in rediscovering the notion of propositional content and so in devising a prop¬
ositional logic, on the basis of a reading of Boethius’s treatise on hypothetical
syllogisms. Abelard saw, as it were, through the incomprehension of Boethius
and LAS to the core of Stoic logic, hidden to them. It is not, therefore, a mat¬
ter strictly speaking of parallel development, but rather of a sort of thoughtful
perspicacity, a logical ability that enabled Abelard to make of the tradition
more than it ostensibly offered him.

Second, Christophe Erismann’s paper shows how, in a limited area, it may
be justified to speak of a Platonic logic. Although Porphyry and many of the
Neoplatonists who interested themselves in logic (especially Boethius, who is
strikingly Porphyrian) allow their Neoplatonism to affect their logical think¬
ing only sporadically, there was a current in the logic practised by Neopla¬
tonists that was more genuinely Neoplatonic, and one medieval Latin thinker,
John Scottus Eriugena—using especially his knowledge of Greek Christian
sources, such as Maximus the Confessor—took it to its limits. True, as Eris-
mann indicates, Eriugenian logic is always shading into metaphysics, but it
begins from Aristotle’s Categories.

Third, Christophe Grellard’s paper brings out the many-rootedness of medi¬
eval logic in a way that Ebbesen’s scheme does not accommodate. Aristotle

134

Introduction / Vivarium 45 (2007) 131-135

himself was working within an already complex philosophical tradition, and
some of the earlier thinkers he attacks would become foundational figures of
later ancient schools. Grellard shows how philosophers like Nicholas of Autre-
court and John Buridan used arguments and counter-arguments about scepti¬
cism in Aristotle’s Prior and Posterior Analytics in their own ways. Some of
these seem to parallel moves made by genuine ancient sceptical philosophers
(and knowledge of Sextus Empiricus, translated but little read, cannot be
ruled out). But what Grellard shows is that even just a close reading of Aristo¬
tle can lead out of the Aristotelian tradition.

Although Ebbesen may be wrong in these ways, there is one, more impor¬
tant, central way in which he is completely right, and in the light of which
even Martins, Erismann’s and Grellard’s powerful points need to be qualified.
Ebbesen’s main point is not so much that we should look at traditions s and t
rather than traditions u and v. It is that we have to be very wary about the idea
of monolithic traditions at all. His view is echoed by Panaccio’s essay. Panaccio
emphasizes the obvious, but often forgotten point, that traditions are a con¬
struct of historians. There were thinkers who had thoughts, and who left
records of them in texts. Panaccio’s own piece, examining the textual chains’
by which theories of mental language were transmitted from antiquity, gives
an example of his method. Other pieces in this collection also illustrate it. It
would be hard, for instance, to think of one of Irene Rosier-Catach’s topics—
the use by Abelard, a twelfth-century nominalist, of the presentation by Priscian,
a sixth-century grammarian who borrows heavily from the Stoics, of Plato’s
theory of Ideas, as transformed (through a tradition going back to Philo
Judaeus) into Ideas in the Mind of God—in terms of anything but individual
texts and their multiform usages. Or again, whilst a contrast between Aristo¬
tle’s exclusion of non-assertive sentences from the realm of logic and the
Stoics’ acceptance of them may set the scene for the controversy Martin Lenz
describes between Abelard and a pupil of Alberic on compositionality, the
reader of this analysis is quickly transported from vague ideas of traditions to
the specifics of a twelfth-century intellectual encounter that is brought to life
with the help of an understanding of contemporary semantics.

Perhaps, then, the main title of this collection, ‘The Many Roots of Medi¬
eval Logic’ hints at a better methodology than the sub-title, which refers to the
Aristotelian and the non-Aristotelian traditions. The best work in the follow¬
ing pages at once brings the variety of ancient logical tradition into focus and
reveals the inadequacy of thinking in terms of traditions. We start to learn
about the past of our subject, it tells us, when we try not to label, but look
carefully at texts and their relations and try to understand the story they tell as

Introduction / Vivarium 45 (2007) 131-135

135

a philosophical one. All that is missing here is an awareness that philosophical
stories, even those about logic, unfold as part of a wider and more brutal
narrative—but that is the task for another book, or a series of them.

The papers which appear here (with the exception of Claude Panaccios) are
based on those given at the 15th European Symposium on Medieval Logic
and Metaphysics, which took place in Cambridge on 1-3 July 2004 at Trinity
College, Cambridge. That conference was generously supported by the British
Academy and Trinity College, Cambridge. My gratitude is due also to the
many participants who, by paying for their own travel and the conference fee,
enabled me to balance the budget; to the forbearance of the contributors
during the long wait for news about publication of the papers, and to those
speakers who agreed to publish elsewhere, so allowing for a manageably-sized
volume. I am delighted that these papers are being published as a special issue
of Vivarium , and I am enormously grateful to Julian Deahl, of Brill, Lodi
Nauta, Editor of Vivarium , his predecessor, Onno Kneepkens and all of the
editorial board, for making this possible. I owe a very special debt to Margaret
Cameron for helping enormously with the unrewarding task of preparing the
volume for publication.

John Marenbon
May 2007

BRILL

Vivarium 45 (2007) 136-152

VIVA

RIUM

www.brill.nl/viv

The Traditions of Ancient Logic-cum-Grammar
in the Middle Ages—What’s the Problem?

Sten Ebbesen

University of Copenhagen

Abstract

Clashes between bits of non-homogeneous theories inherited from antiquity were an
important factor in the formation of medieval theories in logic and grammar, but the
traditional categories of Aristotelianism, Stoicism and Neoplatonism are not quite
adequate to describe the situation. Neoplatonism is almost irrelevant in logic and
grammar, while there might be reasons to introduce a new category, LAS = Late
Ancient Standard, with two branches: (1) logical LAS = Aristotle + Boethius, and (2)
grammatical LAS = Stoics &c. -> Apollonius —* Priscian.

Keywords

Aristotelianism, Stoicism, Neoplatonism, Scholasticism

Medieval scientific practice to a great extent consisted in solving a problem
well-known from latter-day natural science. You have got a couple of mutually
independent theories, each with a field of its own, but also some overlap. How
do you produce a unified theory for the combined field? Being faced with
that challenge is intellectually stimulating and I, among others, believe that
part of the fuel for medieval inventiveness came from the incongruity of some
of their inherited ancient theories, whose incongruity in turn was due to their
different origin.

Thus we are faced with the task of identifying clashes between inherited
theories, demonstrating how those clashes were constructively dealt with,
and tracing the history of the various pieces of theory. For this we need
information about our old thinkers and their works. We also need historio¬
graphical categories, and—as everybody knows—historiographical categories
are in the best case too broad, in the middle case, which is not here the place

DOI: 10.1163/156853407X217687

S. Ebbesen / Vivarium 45 (2007) 136-152

137

of virtue, they are plain useless, and in the worst case they are definitely mis¬
leading. From time to time we must take up our standard categories for evalu¬
ation. So, the belief in the fecundity of clashes between traditions carries with
it problems of its own as well as its own variants of standard problems for
historical research.

How Many Different Ancient “Traditions” Do We Actually
Have to Operate with?

“Traditions” in philosophy are often, and for good reasons, thought of in
terms of schools, the members of which share common tenets and attitudes
thanks to intellectual descent from a common master. Often, members of a
school of thought publicly declare their allegiance to the founder.

Aristotelianism

In this sense it seems unproblematic to pose an Aristotelian tradition in Anti¬
quity, in the Middle Ages, and in the Renaissance, yet it is a commonplace
that “Aristotelianism” is a word of the type about which Aristotle might have
said that it is said in many ways.

In the Western Middle Ages there is one very straightforward sense in which
there is an Aristotelian tradition. People simply studied Aristotle’s works, again
and again, generation after generation, and usually claimed to be in essential
agreement with him. Moreover, none of the Aristotelian works they studied
have been lost in the meantime, so we can read what they read, check on their
quotations, and easily recognize Aristotelian items in their writings.

In another way the Aristotelian tradition is a very elusive entity. Usually a
systematic coherence is assumed to exist between different pieces of doctrine
from the same school. The medieval Aristotelian tradition hardly satisfies the
criterion for being a school in that sense. If “Aristotelian tradition” is under¬
stood as a generic term comprising several specific ways of understanding
Aristotle or several different ways of doing logic and linguistics each one
inspired by Aristotle, but representing different interpretations of his work,
we soon end up with a vaguely defined genus and an unwieldy lot of species
and individuals. We are too well informed about too many Aristotelians!

138

5 . Ebbesen / Vivarium 45 (2007) 136-152

Stoicism and Hellenistic Standard

“Stoicism” is a difficult historiographical category for the opposite reason. We
have so little left of the writings of the Stoic theoreticians that the otherwise
sensible question “Whose Stoicism? Zenos or Chrysippus’ or Diogenes of
Babylon’s or..is for the most part unanswerable, and it is standard proce¬
dure among scholars to take all the ancient reports of what Stoics say and
think to be more or less accurate reports of one coherent philosophical system,
all of whose essential ingredients could once be found in the oral or written
teaching of Chrysippus. Only the assumption that all the pieces of informa¬
tion are information about one system makes it at all possible to try to recon¬
struct any system at all.

Accepting this hypothesis amounts to swallowing a fair-sized camel, but it
must be done. Then we can begin straining at the gnats, and that means doing
old-fashioned Quellenforschung. We would like to take the doctrine of Augus¬
tine’s De dialectica to be Stoic in all essentials—it fills in a big lacuna in our
knowledge about Stoic dialectic, and though the text was not that popular in
the Middle Ages, its being Stoic would provide us with a known channel of
Stoic influence on the medievals, though it only a few of them can be demon¬
strated to have read it. 1 A couple of generations ago there seemed to be little
problem in this. Karl Barwick had convincingly reconstructed some key fea¬
tures of Stoic doctrine of language and in an influential 1957 study he found
De dialectica in accordance with the recipe. As early as 1912 B. Fischer had
plausibly argued that the work reproduced doctrine that Augustine had found
in Varro’s encyclopedic Discipline from the late first century BC. Varro, in
turn, could be relied on to rely on one or more Stoic sources. 2

The serenity of this picture was disturbed by an observation made by Bar¬
wick himself, namely that De dialectica contains at least one feature of Aristo¬
telian origin, namely a distinction between equivocals and univocals. He tried
minimize the damage to his theory by assuming that Varro’s immediate source
was Antiochus of Ascalon who, supposedly, had grafted a small Peripatetic
branch onto a large Stoic trunk. 3 In 1958, however, D. Fehling, threw doubt

1} The work was used by Roger Bacon (who cites it in De signis and in the Compendium studii
theologiae) and by Henry of Ghent: see Rosier (1995b).

2) Barwick (1957), Chapter 1; Fischer (1912).

3) Barwick (1957), 18-22; cf. Ruef (1981), 153 and Pepin (1976b), 68-69. De Rijk (1977, 82-84)
denies that Augustine used a specifically Stoic source. For a “Stoic” reading of De dialectica see
Pinborg (1962), 148-177.

S. Ebbesen / Vivarium 45 (2007) 136-152

139

on whether Varro had actually been as faithful to his mainly Stoic source(s) as
supposed by Barwick. 4

For all that, I stubbornly continue to believe that Varro was Augustine’s
main source, and that Varro’s doctrine was mainly of Stoic inspiration. But
while I am convinced that Stoicism was pervasive in ordinary teaching of
grammar and logic in the centuries round the birth of Christ, I must admit
that in Quintilians Institutio oratoria from the late first century AD there are
Aristotelian features among the ones I know or believe to be Stoic. For instance,
one finds the whole list of the ten categories, and with explicit attribution to
Aristotle, though with a tell-tale Stoicising interpretation of ouoia, which, we
are told, is the concept corresponding to the question an sit , i.e. whether
something is (the case). 5

The situation just described suggests that perhaps it might be useful to
operate with not just an Aristotelian and a Stoic tradition but also with an
entity that we might call Hellenistic Standard, a complex of terminology and
doctrines that cannot be traced back to one great philosopher, and in which
certain Stoic and certain Aristotelian ingredients are readily recognizable, but
which might turn out to be more than the chimerical sum of those heteroge¬
neous parts, and which may have had many more adherents than unadulter¬
ated Stoicism or Aristotelianism—or Epicureanism, for that matter. I am not
aware of any serious attempt to describe Hellenistic Standard Logic, it might
be worth a try.

Neoplatonism and Late Ancient Standard

If we turn our attention to the last centuries of Antiquity, the main logic texts
are commentaries on the Organon composed between ca. 250 and 600. Habit¬
ually, the authors of those commentaries are classified as Neoplatonists, though
it is generally recognized that their Platonism was rather subdued when they
expounded Aristotle. The commentators have enough in common that we
may reasonably call their logic Late Ancient Standard logic (and no competing
sort of logic from the period is known). By contrast, it makes little sense to
speak of a “Neoplatonic Logic”. 6 Very few core ideas are definitely Neopla¬
tonic, and still fewer are both decidedly Neoplatonic and relevant to logic,
even taking ‘logic’ in a very wide sense.

4) Fehling (1958).

5) Quintilian, Institutio oratoria 3.6.23-24.

6) In spite of some brilliant papers on the topic by A.C. Lloyd, notably Lloyd (1956).

140

S. Ebbesen / Vivarium 45 (2007) 136-152

One pervasive idea in Neoplatonism is that a millet seed with its enormous
potential is ever so much better than a finished plant. This should be remem¬
bered when we look at an apparent convergence between Stoicism and Pla¬
tonism: both end in their ontological analysis with a non-descript this or
something or there is or one. But the Stoic this is the bare bones of a living
being, the Neoplatonist one is the beast itself. One variant of the Neoplatonist
story about the millet seed is the story about the way from higher reality
to language and back again. When Proclus—in non-Aristotelian contexts—
speaks about symbols, it is easy to see that what he says is very un-Aristotelian,
and we can happily apply the label Neo-Platonic. 7

But when Boethius speaks about signification in his second commentary on
the Perihermeneias , all we can say is: some of the ideas involved must have
originated in a non-Aristotelian context of the sort we call Neo-Platonic. 8
And when the same Boethius in his commentary on chapter 1 of the Catego¬
ries presents what is essentially Porphyry’s account of equivocals, univocals
and denominatives, including the mention of the last two items of old Speu-
sippus’ system, multivocals, that is, and diversivocals—and including the sub¬
division of equivocals into equivocals by chance and by purpose as well as
a discussion of the possible inclusion of metaphor under equivocation—
when Boethius does all that, he is indisputably transmitting the doctrine of a
first-rank Neoplatonist, namely Porphyry—but is the doctrine in question
Neoplatonic in any interesting sense? Not as far as I can see, though a small
Platonic twist may be detected here or there. For all we know, Porphyry him¬
self could have taken over most of his scholium from Alexander of Aphrodi-
sias, who taught Peripateticism for a living in the years round AD 200, and
at least one ingredient, the Speusippean classification, is recorded to have
been already in the work of the older Peripatetic Boethos of Sidon (probably
late 1st c. BC). 9

Or what about the notion of two impositions of names? The medievals
learned it from Boethius, and his source was Porphyry. 10 But it is hard to see

_ 7 - ] Higher reality to language and back again: see Ebbesen (1987). For Proclus on symbols, see

Bonfiglioli and Marmos chapter in this volume.

8) For the passage in question (Boethius 1880, 20-23) see Magee (1989), 76fF., and the reviews
of this book by Ebbesen (1991) and by Mann (1993).

9) Boethius (1847), 163D-168D; Porphyry, Commentary on the Categories , CAG 4.1, 63-70.
For Boethos, see Simplicius, Cat., CAG 8: 38.19ff.

10) Boethius (1847), 159A-C; Porphyry, Cat., CAG 4.1: 37-58. More discussion of Porphyry’s
theory of imposition in Ebbesen (1981).

S. Ebbesen / Vivarium 45 (2007) 136-152

141

anything particularly Neoplatonic about the double imposition, though it is
unknown from any source earlier than Porphyry. To my mind, the most prob¬
able place of origin is Stoicism. The Stoics had developed the Platonic notion
of imposition of names, and in their philosophy there was good reason to
distinguish between a primary imposition of names on ordinary things and a
secondary imposition of names that group words by function. 11 It is not really
important whether you believe me on this or not, because we are dealing with
a piece of theory that, however interesting it may be, is fairly neutral when it
comes to Stoicism versus Neoplatonism. We can safely put the two imposi¬
tions down as an item of Late Ancient Standard rather than as Neoplatonic.

A final example. Boethius’ introduction to categorical syllogisms depends
on a work by Porphyry, 12 whose own sources are difficult to track down. Again,
it is hard to locate anything specifically Neo-Platonic in the work. We do not
have much material for comparison, but Boethius and Porphyry may well
have followed the late ancient mainstream tradition in syllogistic, itself quite
possibly a continuation of one branch—a Peripatetically oriented one—of
Hellenistic Standard. Boethius’ work on hypothetical syllogisms was long a
major source of wonder. Nothing was like it. But recently Susanne Bobzien 13
discovered that a brief Greek text printed by Waiz in the preface to his edition
of the Organon offers a rather close parallel to one part, at least, of the enig¬
matic Boethian work. The Greek text cannot be dated, but chances are that it
reflects a Late Ancient Standard treatment of the matter that we had so far
only heard about from Boethius.

So, to the medievals Boethius was a rich source for Late Ancient Standard
logic, not for Neoplatonic logic, because there never was such a thing.

Neither was there ever a Neoplatonic tradition in grammar. The only
ancient school ever to have a grammar of its own was the Stoics, and impor¬
tant features of their system had been absorbed into Apollonian-Priscianic
grammar without being thoroughly harmonized with either Aristotelian or
(Neo)platonic doctrine, both of which have also left traces in Priscian, who

n) Ebbesen (2005a).

12) Boethius’ Introductio in syllogismos categoricos exists in two variants, the second one incomplete.
The passage that reveals the source occurs in book II of the first version C'De syllogismo categorico ”),
PL 64: 813C, here quoted from a forthcoming edition by Christina Thomsen Thornqvist, Uni¬
versity of Gothenburg: “secunda uero figura habet sub se quattuor modos, tertia uero auctore
Aristotele sex. Addunt etiam alii unum, sicut ipse Porphyrius superiores scilicet sequens”.

13) Bobzien (2002).

142

S. Ebbesen / Vivarium 45 (2007) 136-152

was to be the only ancient grammarian with any great impact on scholastic
linguistic thought—Donatus, though used in every cathedral school, was to
jejune to offer much inspiration except for situations in which he disagreed
with Priscian. As Boethius’ logical writings are better classified under Late
Ancient Standard than under a school name, so it is less misleading to think
of Priscian as a representative of Late Ancient Standard grammar than as a
representative of some philosophical school.

Vaguely Neoplatonic hierarchical approaches to problems were known to
the medievals from a large number of ancient non-logical and non-grammarial
sources, and many schoolmen made lavish use of the notion of worthiness: in
any set of things, including very abstract ones, some things are worthier or
nobler —digniores or nobiliores —than others. According to Ros(celin Junior,
early 12th c.), the category of substance is worthier than the remaining ones, 14
and according to Anonymus Aurelianensis II (second half of 12th c.), the art
of demonstration is worthier than the other parts of logic. 15 This was still true
to a scholar who worked a century later (1270s), and a near-contemporary of
his in very Platonist-sounding words declared that among the various sorts of
multiplicity that cause the Aristotelian fallacies in speech, equivocation most
genuinely deserves the name because it “takes part in the name in a better and
nobler way” than the remaining five. 16

The remarkable thing is that such arguments from dignity and nobility are
absent from Boethius’ commentaries. True, he does explain that substance is
naturally prior to the accidental categories, but this is good Aristotelian termi-

14) Ros<cellinus?>, Glossulae Categoriarum , ms. Archivio Capitolare della Basilica Ambrosiana,
M2: 3vA, on Arist., Cat. 5 (transcribed by Iwakuma, slightly modified by me) “Primitus de
substantia agit— namely Aristotle —eo videlicet quod substantia omnium aliorum praedicamen-
torum rerum subsistamentum sit, dignius vero est quod [fort, quia cod.} substentat eo quod
substentatur. Sed etiam praedicamentum dignius ceteris est eo quod unumquodque vocabulum
huius praedicamenti aliquam existentiam continere [habere] habeat per appellationem et nullum
de aliis hoc habeat. Dignius etiam dicitur praedicamentum hoc modo significandi, quia vocabula
istius praedicamenti non denotant res suas tanquam assistentes, sed aliorum vocabulorum prae-
dicamentorum vocabula res suas tanquam in subiecto. Sessio enim eundem intellectum facit
quern sedens, sic et alia. Et ita dignius est hoc praedicamentum ceteris. Et ideo de hoc primitus
[quam] utpote de digniori.” (I propose to call the author Roscellin Junior to distinguish him
from his more famous predecessor). Very similar expressions occur in other 12th-c. commentar¬
ies on this passage.

15) Ebbesen (1976), 37.

,6) Anonymus C, in Incerti auctores (1977), qu. 809: Anonymus SF, in Incerti auctores (1977),
Quaest. SE , qu. 30.

S. Ebbesen / Vivarium 45 (2007) 136-152

143

nology, he does not employ value-terms, 17 and so he cannot be held responsi¬
ble for the “Neoplatonic” reliance on arguments a dignitate evidenced from
the very early days of Western scholasticism. In later scholasticism Aristotle
himself may have contributed to the popularity of arguments from nobility.
When his Metaphysics began to be commonly read in the 13th c. it may have
reinforced an already existing tendency because in several passages the transla¬
tions had dignior, nobilior, nobilissimus for djueivcovPeXricov, dpicrrog. 18 Some
authors from the 1270s simply identify natural priority with nobility. 19

While such hierarchical thinking owed much of its popularity to ancient
(Neo)Platonists, as a “tradition” or as symptom of a tradition it is too diffuse
to be of much use. And, to be fair to the medieval logicians and grammarians,
for the most part the use they made of arguments a digniori et nobiliori was
rather innocuous, since they were mainly applied to ultimately inconsequen¬
tial matters such as the right order of presentation of items. The two authors
from the 1270s whose use of such arguments was referred to above, are ada¬
mant that standing higher or lower in terms of value is irrelevant to whether
two things are species of the same genus and, consequently, whether a com¬
mon designation applies to them univocally or equivocally. They do not want
to get entangled in the theologians’ confusion of ordinary prius—posterius
analogy with nobilius—vilius analogy.

Conclusions about Ancient Traditions

However one divides Hellenistic logic into schools or traditions, virtually
nothing from that period exercised a direct influence on the Middle Ages.
Small pieces of theory that might perhaps best be assigned to Hellenistic Stan¬
dard were known from Cicero (notably from his Topics ), and little else. Broadly
speaking, ideas of Hellenistic origin were accessible to the medievals only
as absorbed into Late Ancient Standard. There is an important difference,
however, between logic and grammar, between Boethius and Priscian. While
nobody doubts that fragments of Stoic lore float around in Boethius’ works,
they certainly are not dominant, and moreover it is often very hard to certify
the Stoic origin of a suspected fragment. By contrast, in Priscian Stoic ingre¬
dients are both obvious and more dominating.

17) Boethius (1847), 182A.

18) See indices in the Aristoteles Latinus editions.

19) Anonymus C, in Incerti auctores (1977), qu. 809. Anonymus SF, in Incerti auctores (1977),
Quaest. SE, qu. 57.

144

S. Ebbesen / Vivarium 45 (2007) 136-152

General Neoplatonic ideas, not specifically oriented towards logic or gram¬
mar, were transmitted to the medievals by any number of sources—Augustine,
Pseudo-Dionysius, Boethius...—but because of their ubiquity and lack of
specifically logico-grammatical content it is hard to pinpoint a “Neoplatonic
tradition” in the field under investigation.

At least in early scholasticism, the only two both important and recogniz¬
able competing “ancient traditions” are: (1) the Aristotle-based logical one
represented by Boethius and Aristotle himself, (2) the Apollonio-Priscianic
grammatical tradition with its strong Stoic traits.

Parallel Developments

Since the medievals possessed Aristotle’s logic, no problem is caused by the
occurrence in their works of ideas that look Aristotelian. We do, however,
frequently have a problem with ideas that look Stoic. If they cannot be traced
back to Priscian, we are often at a loss to explain how they could possibly have
been transmitted through the centuries.

An egregious example is that of the medieval dictum or enuntiabile. The
concept is there from early in the 12th century, at least, 20 and it has strong
similarities to the Stoic Xektov, the incorporeal quasi-thing which is the inter¬
face between sign, thought and genuine, corporeal things. We even find enun-
ciables listed with such items as time, place and the void as examples of
quasi-things—just as the Stoics are known to have listed sayables, time, place
and the void as incorporeal somethings. 21

Much energy has been spent on finding the ways of transmission of the
notion of Xektov from the Stoa to the 12th century, but no convincing route
has been found. The traces of the Xektov in Senecas Letters and Augustine’s De
dialectica cannot explain what happened in the twelfth century, nor are those
ancient authors referred to on this matter. I think we must admit that the
similarity of the dictum or enuntiabile to the Xektov is due to parallel develop¬
ment: similar conditions produce similar results. The concept of parallel devel-
opement is one I have borrowed from evolutionary biology. An American and
a European bird may have very similar properties while being genetically
worlds apart, similar environmental challenges having been met with similar
responses.

20) The fundamental study is Nuchelmans (1973). Among several later studies I may mention
Iwakuma (1997) and Kneepkens (1997).

21) See Ebbesen (2004), 113.

S. Ebbesen / Vivarium 45 (2007) 136-152

145

If this is accepted we must ask ourselves what similarities in the initial
conditions produced similar effects in the two cases. I would point to two
factors:

First, both the Stoics and the medievals thought in terms of the pair signa —
significata , also known as signa — res , the Stoics inspired, no doubt, by both
Plato and Aristotle, and the medievals by both Aristotle and the Stoics. The
fundamentality of this pair had been stressed by the Stoics and that part of
their doctrine was available to the medievals in several sources.

Second, both Stoics and medievals were occupied with keeping their ontol¬
ogies tidy. The Stoic system appears to have been fairly simple: there are genu¬
ine existing things, all bodily; there are secondary incorporeal things such as
happenings which, though caused by bodies, cannot be eliminated from our
analysis of reality; and finally there are eliminable mental constructs such as
Platonic or Aristotelian forms. The doctrine of Xektcx was, not least, designed
to keep the ontology tidy.

The medievals were much more friendly towards incorporeals, but they too
wanted to keep their ontology tidy, and in the twelfth century, if no before,
people became acutely aware of the fact that philosophy operates and needs to
operate with items that have no place in the Aristotelian system of categories.
There was, of course, an old worry about whether the categories are at all rel¬
evant to God. But now it also became obvious that other items must be
exempted from categorial status—enunciables were just one such item, along
with ens and aliquid , among others. 22

In short, the picture, as I see it is this: the Stoics had left a tradition for wor¬
rying about the ontological status of things that are neither bodies nor spiri¬
tual doublegangers of bodies, but about which we seem to be speaking all
the time. They also had left a tradition for thinking in terms of signs and
significata. It was not strange, then, that twelfth-century thinkers should start
to theorize about “that which I state when I make a statement”. They were in
need of a concept to do the job of a Xektov, there was an empty space, as it
were, in their theories waiting to be filled in. To use the language of evolution¬
ary biology, we might say that a theory with a Afxrdv-like ingredient would be
more likely to survive in the struggle with opposing theories than one without
such an ingredient—in fact, the twelfth century was characterized by fero¬
cious competition between the masters who taught philosophy.

22) Cf. S. Ebbesen, (2005b), and L. Valentes contribution to the present collection.

146

S. Ebbesen / Vivarium 45 (2007) 136-152

The Importance for Medieval Developments of Clashes between
Non-homogeneous Parts of Their Inherited Intellectual Baggage

I shall now discuss a couple of cases in which combinations of and clashes
between non-homogeneous parts of the inherited intellectual baggage seem to
have spurred medieval theoretical inventiveness—though of course not always
with tenable results.

First Example: Transitivity

The Stoics were, as far as we can see, the first to elaborate a reasonably complex
theory of what makes a sentence complete and congruent. They distinguished
between, on the one hand, one-place predicates such as walking or ‘snoring
that only require a subject to produce a complete sentence, and, on the other
hand, such two-place predicates as “loving” or “beating” that also require an
object.

“The man is walking” would be a complete sentence describing some thing
with some property in a certain state, i.e. a thing falling under the three first
Stoic categories something—of some quality—in some state. “The man is beating
his dog” would be a complete sentence describing a thing with some property
in a certain state relatively to something’, i.e. a thing falling under all four
Stoic categories, the last of which was in some state relatively to something.

In Priscian, and so to the medievals, 23 “to beat” is a transitive verb, and the
notion of transition (peidpaaic;) is also found in Priscian’s main source of
inspiration, Apollonius Dyscolus. There is no evidence for it before then, but
there is good reason to think it is a Stoic one. The basic idea is clear enough:
such verbs describe happenings which originate in the subject person and pass
on to the object person.

As Irene Rosier-Catach has stressed on numerous occasions, in about the
second quarter of the thirteenth century, Latin grammar was massively invaded
by the terminology of motion from Aristotle’s Physics. Since the metaphor of
transition is drawn from motion, transitivity also began to be described in
terms derived from the Physics.

And once you are in the world of the Physics you are in the world of the four
Aristotelian causes. An example of ingenious grafting of Aristotelian branches
on a Stoic trunk occurs in a 13th-century grammatical analysis of the phrase
legendum est. This is a complete sentence, but difficult to analyze if you do not

23) On Priscian as the medievals source for the notion of transitivity, see Kneepkens (1995), 245.

S. Ebbesen / Vivarium 45 (2007) 136-152

147

possess the concept of a monovalent predicator. Legendum was thought to be
an accusative, hence a nominative, such as dignum or debitum must be under¬
stood. But then the construction must be transitive, and legendum must be the
terminus or finis of the transit. What sort of end? A final cause, of course!—
Neckbreaking, but inventive! 24

Second Example: What Is a “vox?

Aristotle, in Categories 6, 4b23, uses Xoyoq / oratio to exemplify discrete quan¬
tity and explains (4b33-35) that doing so is justified because Xoyoq / oratio is
measured by long and short syllables. At the very beginning of his Institutiones
grammaticae (1.1.1) Priscian—following a Stoic materialist understanding of
vox —claims that vox is air that has been made to oscillate (aer ictus) , and the
very same claim occurs at the beginning of Donatus’ Ars maior. Since an oratio
must be a vox , and air must be a substance (in Aristotelian terms), medieval
scholars soon spotted a problem of how to reconcile the two authorities. I will
not go into details; suffice it to say that texts from the early 12th century, pre¬
sumably from the surroundings of William of Champeaux, suggest that the
solution may be that the two ancient authors viewed vox from two different
points of view: Priscian was thinking of the material substrate of linguistic
items, chunks of oscillating air, we might say. Aristotle was thinking of the
tenores , alias accentus , suprasegmental phonemic ingredients that differentiate
such chunks of air from similar chunks and make them identifiable to a
listener as probably carrying a message from a fellow human being. This is
breathtaking! Via puzzling about a disagreement between two authorities early
12th-century men took great strides on the way to a reasonable ontology
of words. 25

Third Example: Substance , Quality and “res ”

According to Donatus a noun signifies some body or thing in a common or in
a private way. 26 The thing in case must be incorporeal for the disjunction to

24) Iohannes(?), Sicut dicit Remigius S13 (Legendum est musas), ms Paris BNF lat. 16618:
76rA.Text in Rosier-Catach and Ebbesen (2004).

25) I gratefully acknowledge having had access to unpublished work by Margaret Cameron
(Hunter College, CUNY), who has gathered relevant sources. For the kind of problem involved
in this discussion, see Ebbesen (2003b).

26) Donatus, Ars maior II.2, ed. Holz 1981: 604 : “Nomen est pars orationis cum casu corpus aut
rem proprie communiterque significans”.

148

S. Ebbesen / Vivarium 45 (2007) 136-152

make sense, and though this definition may not have been the “official” Stoic
definition of the noun, it relies on Stoic ontology which comprises first-class
beings called bodies and second-class incorporeal quasi-beings called happen¬
ings, npaypaza , not very happily translated as res by Latin grammarians.
According to Priscian, a noun is a part of speech which distributes to each of
its underlying bodies or res a common or a private quality. 27 Again this is
nowhere attested as a Stoic definition, whereas several sources claim that the
Stoics did not treat common and proper names as one part of speeech. Yet, it
reeks of Stoicism. Not only do we find the division of things into bodies and
happenings, we also find the Stoic pair of common and private qualities, and
on top of all the definition is etymologizing in a totally Stoic manner: a noun,
an ovopa, vepeu distributes, a quality. 28

The same basic idea is expressed in the short Priscianic formula most often
used by the medievals: a noun signifies a substance and a quality. 29 The sub¬
stance is not an Aristotelian one but a Stoic ‘something such as can be signified
by means of the demonstrative ‘this’ or the article ‘the’, which contains no
information about what sort of thing the speaker is pointing at. A noun is
more informative, signifying something of some sort, a noiov , a quale or quale
quid —which of course is reminiscent of Aristotle’s remark about secondary
substance in Categories 5.3b 15-16, and here as there quality has to be inter¬
preted as ‘property’ in general.

So, from the ancient definitions of the noun the medievals had access to
some fundamental un-Aristotelian features of Stoic ontology and semantics.
They did not, however, receive it with too much explanation, so there was
room for interpretation, including complete misunderstanding.

John Scottus Eriugena, seeing that qualitas must have a wider sense than
the category of quality, but failing to recognize the radical departure from the
Aristotelian lore of categories, takes the word to stand for all nine sorts of
accident. 30

27) Priscian, Inst. 2.5.22, GL 2: 56-57: “Nomen est pars orationis quae unicuique subiectorum
corporum seu rerum communem vel propriam qualitatem distribuit”. See parallel Greek texts in
Scheider 1910, Grammatici Graeci II.3: 38.

28) Possibly inspired by Plato, Cratylus 430-431 where diavepeiv is used about assigning names,
ovojuarcc, to things. Notice that subiectis = vnoKeifievcov seems to introduce the Stoic category of

V7tOK£lfl£VOV.

29) Priscian, Inst. 2.4.18, GL 2: 55 “Proprium est nominis substantiam et qualitatem significare”.
In the Middle Ages usually quoted with cum qualitate substituted for et qualitatem.

30) Eriugena, in his glosses on Priscian, ed. Luhtala 2000a: 158 “Sciendumque quod in hoc loco
Priscianus qualitatem posuit pro omnibus accidentibus.” Cf. Anon. Paris lat. 7505, Luhtala
(2000a), 170.

S. Ebbesen / Vivarium 45 (2007) 136-152

149

Peter Helias presents three main interpretations of the claim that a noun
signifies a substance with a quality: 31

(1) A Porretanean interpretation. Substance is id quod , quality id quo.

(2) An interpretation that in the main is that of William of Conches. The
definition is only valid for the model case of nouns signifying real substances,
and they signify a substance in such a way that they delimit a common or
private quality around it —Sed que significant substantiam ita earn significant
quod qualitatem circa earn terminant vel communem [...] vel propriam —.
Notice the tacit reference to Categories 5.3b20 genus autem et species circa sub¬
stantiam qualitatem determinant.

This view Aristotelizes the definition as much as possible, but at the same time
introduces the non-Aristotelian notion of such private qualities as Platos Pla-
tonicity. The grammarians had kept this Stoic notion alive, well helped by the
fact that the basic analysis of things as somethings-with-a-quality was also
acceptable to ancient Platonists, some of whom could also accept ideas of
individuals and most or all of whom could at least accept individual qualities
understood as bundles of properties, each bundle uniquely identifying one
individual. In fact, when mentioning Platonicity, Peter Helias refers to
Boethius on the Perihermeneias . 32

Incidentally, holders of the second opinion also proposed that, strictly speak¬
ing, a substantive noun signifies only the relevant quality, whereas it names the
substances that bear that quality. This is a good example of constructive think¬
ing about a difficult authoritative text.

(3) The third opinion—traceable back to Thierry of Chartres—also takes “sub¬
stance” to be Aristotelian substance, though strictly speaking only prototypical
nouns signify substance. The quality in case is taken to be the grammatical
quality of signifying in a common or a private way. 33

The third opinion has erased virtually all traces of Stoicism, but not the two
first. A watered-down variant of the Porretanean position became widely

31) Peter Helias (1993), 189fF. Cf. De Rijk’s treatment of the passage in (1962, 1967), II.1,
229-234.

32) Peter Helias (1993), 191. The reference is to Boethius (1880), 137. Cf. the corresponding
text ofWilliam of Conches in Fredborg (1981), 30-31 and De Rijk (1962, 1967), II.1, 223.

33) Peter Helias (1993), 194. Same opinion in Anon. Comm. Prise., ms Wien, ONB, lat. 2486,
quoted (1962, 1967), II. 1, 242. For Thierry, see the quotation from CIMAGL 21 below.

150

S. Ebbesen / Vivarium 45 (2007) 136-152

accepted in the late 12th c., and is for instance expressed by another commen¬
tator on Priscian who says that homo signifies a quality, namely humanity, as
adiacens , which it apposes to the supposit (“subject” in later terminology)
when it, homo , that is, stands in appositum- position (“predicate position” in
modern parlance). 34 Here a substantial form is as neatly separated from its
bearer as are accidental forms, and this is used to explain predication in quid
along the same lines as accidental predication. The same author, and several
others, including a commentator on the Categories , also take the grammarians
substance to be suppositum locutioni , that about which one speaks. 35

This is not the same as retrieving the genuinely Stoic notion of a urcoKEijievov,
a thing that we have mentally stripped of all characteristics except being there,
but wouldn’t someone get the idea?

Another twelfth-century view was that the noun’s signifying substance and
quality means that it signifies an essentia and some property “For”, one source
says, “Priscian does not use “substance” in the same sense as does Boethius.
Priscian calls any essentia a substance”. 36 Whatever exactly such a 12th-century
essentia is, it is something which transcends the Aristotelian categories and is
not really at home in Aristotle’s philosophical castle.

The late thirteenth century was in many ways the climax of Aristotelianism
at the universities, but if we look at what modist grammarians did with the
definition of the noun, we find something that is much rather Stoic than Aris¬
totelian. According to Martin of Dacia the essential general mode of signify¬
ing of the noun is that of signifying per modum habitus et quietis etper modum
determinatae apprehensionis. This is a composite mode, in which the modus
habitus et quietis has the role of matter and the modus determinatae apprehen¬
sionis has the role of form. And this, he explains, is what the ancients used to
express in the formula “every nouns signifies a substance with a quality”, which
must be understood “modally”, he adds, so that it means per modum substan¬
tiae plus per modum qualitatis. In other words, he interprets the formula as

34) Glosa “Promisimus ”, ed. in Fredborg (1999), 149: “‘Homo’ autem significat qualitatem ut
adiacentem, sc. humanitatem, et earn supposito apponit quando est in apposito, ut ‘Socrates est
homo’.” My punctuation.

35) Glosa “Promisimus ”, ed. in Fredborg (1999), 149: “que substantiam significant id est sup¬
positum locutioni”. Anon. D’Orvillensis, Cat., ed. Ebbesen (1999), 274. Cf. Anon., Tractatus
glosarum Prisciani , ed. Fredborg 1977: 43 “Aliter enim dicit magister Thedricus: significat sub¬
stantiam et qualitatem i.e. suppositum de quo loquimur communiter vel proprie ita quod con-
veniat uni vel pluribus.”

36) Anon., Comm. Prise., ms Wien, ONB, lat. 2486, quoted in De Rijk (1962, 1967), II. 1, 241.

S. Ebbesen / Vivarium, 45 (2007) 136-152

151

claiming that a noun signifies in such a way as to suit a prototype case in
which its significate may be analysed as something stable grasped in a determi¬
nate way. 37 We are very close to the Stoics here! The same may be said about
the way other modist authors define the noun—Thomas of Erfurt and Michael
of Marbais do not disagree significantly from Martin on this issue, and Radul-
phus Brito comes even closer to the Stoics by saying that the noun signifies per
modum essentiae determinatae , adding that the underlying mode of being is esse
formale in quolibet ente —be it extramental or mental, privative or positive. 38

A very complicated case of clash between Stoic and non-Stoic terminology
regards the npaypa, that we have already met in the description of a noun as a
word that signifies a body or a xpaypa. Elsewhere in this volume Anne Gron-
deux discusses what the Latins did with the res they got for a npaypa, so I can
restrict myself to pointing to some long-term developments. In the definitions
of the noun the Latins got a res that must be incorporeal. In Priscian Minor
they also learned that infinitives signify the very res that the verb contains, for
to run is running, to write is writing, and to read is reading: “significat autem
infinitum ipsam rem quam continet verbum: currere enim est cursus et scribere
scriptura et legere lectio”. 39

If the medievals had not understood too well what type of incorporeals
must be meant by res in the definitions of the noun, this passage could help
them. The res verbi became a well-established notion of medieval grammar,
and as was clear from the Priscianic passage it must in itself be above the noun/
verb distinction.

Within a theory of predication that was, or at least pretended to be Aristote¬
lian, John Pagus and some other 13th-c. writers employed the un-Aristotelian
res verbi to explain how ampliation and restriction works. 40 Interesting as this
may be, probably the most interesting result of having such un-Aristotelian res
around resulted from its marriage with the Avicennian common nature. The
res that words signify according to late-13th.-c. modists are neither Aristote¬
lian nor Stoic nor, I think, Avicennian. 41 But their struggle with the conflicting

37) Martin of Dacia (1961), 10-11

38) Thomas of Erfurt (1972), 154; Michael of Marbasio, Summa (1995), 19-20; Radulphus
Brito (1980) 192-193.

39) Priscian, Institutiones grammaticae 18.4.43, GL 3, 226. My punctuation.

40) Ebbesen and Rosier-Catach (2000), 88-91.

41) For the role of Avicennas common nature, see Pinborg (1967), 44-45.

152

S. Ebbesen / Vivarium 45 (2007) 136-152

notions of substance, quality and res in their sources, had left the medievals
with highly abstract notions of res , liberated from the embrace of tangible
three-dimensional objects, and ready for the link-up with common natures.

Conclusions

I hope to have made plausible what many scholars believe anyhow, namely
that the clashes between bits of non-homogeneous theories inherited from
antiquity were, indeed, an important factor in the formation of medieval
theories. And while not rejecting the traditional classifications into Aristote¬
lian, Stoic and Neoplatonic, I have given some reasons for being careful, sug¬
gesting that we might need at least two new categories: Hellenistic Standard
and LAS = Late Ancient Standard, the first of which is of scant relevance for
the Middle Ages, while the latter cleaves into two: logical LAS = Aristotle +
Boethius, and grammatical LAS = Stoics & c. —> Apollonius —> Priscian.
Neoplatonism is a concept to be invoked only on rare occasions in the history
of logic and grammar.

BRILL

Vivarium 45 (2007) 153-168

RIUM

www.brill.nl/viv

Denying Conditionals: Abaelard and the Failure of
Boethius’ Account of the Hypothetical Syllogism

Christopher J. Martin

Auckland University

Abstract

Boethius’ treatise De Hypotheticis Syllogismis provided twelfth-century philosophers
with an introduction to the logic of conditional and disjunctive sentences but this work
is the only part of the logica vetus which is no longer studied in the twelfth century. In
this paper I investigate why interest in Boethius acount of hypothetical syllogisms fell
off so quickly. I argue that Boethius’ account of compound sentences is not an account
of propositions and once a proper notion of propositionality is available the argument
forms accepted by Boethius are seen to be incoherent. It was Peter Abaelard who first
understood the nature of propositionality and propositional connectives and used this
to criticise Boethius’ claims in De Hypothetics Syllogismis . In place Boethius’ confusion
Abaelard offered a simple and correct account of the hypothetical syllogism.

Keywords

Boethius, Abaelard, hypothetical syllogism, negation, conditional

The statues promulgated in 1252 1 and 1255 2 to regularise teaching in the
Faculty of Arts at the University of Paris show that logic, as it should, had a
central place in the training in Arts. Students were required according to the
statute of 1255 to work their way through: ‘the old logic, that is, the book of
Porphyry, the Predicaments , the Peri her menias, and the De Divisione and Topics
by Boethius excluding the fourth book’. 3 The old logic according to the Parisian

0 Denifle (1889-1897), I, n. 201.

2) Denifle (1889-1897), I, n. 246.

3) Denifle (1889-1897), I, n. 246: ‘Veterem logicam, videlicet librum Porfirii, predicamento-
rum, periermenias, divisionum et thopicorum Boecii, excepto quarto...’ Lectures on this mate¬
rial lasted for two terms beginning on the 1st of October, the Feast Saint Remigius, and ending
on the Feast of the Annunciaiton, March 25th.

DOI: 10.1163/156853407X217696

154

C. ]. Martin / Vivarium 45 (2007) 153-168

masters thus consisted of works already available at the beginning of the
twelfth century with one very notable exception: Boethius’ De Hypotheticis
Syllogismis {DHS) is nowhere to be found in the curriculum. 4

Written a century and half before the statues were drawn up, both the
Dialectica of Garlandus Compotista and that of Peter Abaelard are divided
into treatises corresponding to the then accepted canon of logical works. In
each of them the hypothetical syllogism receives a very extensive treatment:
one third of the whole in Garland’s case and about a fifth of what remains of
Abaelard’s book are taken up with its treatment. For Abaelard, furthermore,
the discussion of the topics is inseparable from that of the hypothetical syllo¬
gism since the topics provide the conditional propositions to be used in such
arguments. The treatment of these two together takes up 350 of the 550 or so
pages which survive of Abaelard’s Dialectica.

The present paper attempts to address the question of why interest in
Boethius’ book and in the hypothetical syllogism in general did not survive into
the thirteenth century. The answer, I think, lies at least in part in the incompat¬
ible accounts of what constitutes the opposite of a given conditional found by
twelfth century logicians in the material which they inherited from antiquity.

Boethius

In introducing his study of the hypothetical syllogism Boethius complains
that the subject has been discussed only briefly by a few Greek writers and not
at all in Latin. 5 There is no reason, I think, to doubt that he is accurately
reporting the information available to him. Of the Greeks he mentions only
Theophrastus and Eudemus and he shows no knowledge at all of Stoic discus¬
sions of hypothetical syllogisms. 6 Boethius goes on, on the other hand, to

4) The earliest reference to the logica vetus is in the statute of 1215 ( Chartularium n. 20):
‘... legant libros Aristotelis de dialectica tarn de veteri quam de nova in scolis ordinarie et non ad
cursum’. However, no account is given there of the works included in it.

5) Boethius, De Hypotheticis Syllogismis (DHS) 1.1.3 (Boethius, 1969): ‘Quod igitur apud scrip-
tores quidem Graecos perquam rarissimos strictim atque confuse, apud Latinos uero nullos rep-
peri, id tuae scientiae dedicatum noster etsi diuturnus, coepti tamen efficax labor excoluit.’

6) DHS 1.1.3: ‘...de hypotheticis syllogismis saepe quaerebas, in quibus nihil est ab Aristotele
conscriptum. Theophrastus uero, uir omnis doctrinae capax, rerum tantum summas exsequitur;
Eudemus latiorem docendi graditur uiam, sed ita ut ueluti quaedam seminaria sparsisse, nullum
tamen frugis uideatur extulisse prouentum.’ His discussion of the Stoic indemonstrables in his
commentary on Ciceros Topica , In Ciceronis Topica , is based entirely on the information given
by Cicero. See Martin (1991), 277-301.

C. J. Martin / Vivarium 45 (2007) 153-168

155

characterise his own work as an elucidation and development of what has been
already discussed by his predecessors. 7 The fact that he had only a little to go
on and perhaps drew on different sources may well explain some of the pecu¬
liarities of this development. 8

I have argued elsewhere that although Boethius certainly provides a set of
inference rules for conditional and disjunctive propositions in DHS and so a
logic for at least some such propositions, he does not provide us with a propo¬
sitional logic, or even a fragment of a logic, as we understand it today. 9 This is
because he does not possess our concept of propositional content and so does
not understand propositional form and propositional operations as we do.

I will not argue the point again here but an example of the difference is that
for us any uniform substitution instance of the schema £ if p , then q,p\ therefore q
is a valid argument no matter how complicated the propositional contents we
might substitute for 'p and ‘ q\ For Boethius, on the other hand, modusponens
has to be defined separately for each acceptable variation of the major premiss.
He does, it is true, allow unlimited substitution but the substitution is of general
terms for the term variables which occur in each of his argument schemata.

Boethius, indeed, attempts to calculate just how many different forms of
conditional proposition there are. He distinguishes simple conditionals with a
categorical antecedent and consequent from composite conditionals in which
either the antecedent, or the consequent, or both is a simple conditional.

In addition he designates certain pairs of simple conditionals as mediate’
conditional propositions. These pairs are of three different kinds. In the first the
consequent of the first conditional and the antecedent of the second are the
same, for example, 'si esta ,, est b, si est b } estc. 10 In the second the antecedent of
the second conditional is the negation of the antecedent of the first. In the third
the consequent of the second is the negation of the consequent of the first.

7) DHS 1.1.4: ‘Nos igitur, quantum ingenii uiribus et amicitiae tuae studio sufficimus, quae ab
illis uel dicta breuiter uel funditus omissa sunt, elucidanda diligenter et subtiliter persequenda
suscepimus...’

8) Susanne Bobzien discusses a Greek scholium to the Prior Analytics that contains an account
of various kinds of hypothetical syllogism in which she claims to find close parallels with DHS.
The scholium distinguishes (1) syllogisms with a simple conditional and a categorical as prem¬
isses, (2) syllogisms with a simple disjunction and a categorical as premisses, (3) syllogisms with
two conditional premisses, i.e. wholly hypothetical syllogisms, though they are not so named in
the text. Syllogisms of type (3) are divided further into three figures corresponding to Boethius’
figures for syllogisms with a ‘mediate’ conditional as their major premiss. See Bobzien (2002,
286-300).

9) Martin (1991).

10) DHS 2.9.1.

156

C. J. Martin / Vivarium 45 (2007) 153-168

These are all the kinds of conditional proposition that Boethius recognises.
His calculation is complicated by the two qualities and the five modalities that
he allows to simple categoricals, 11 but in the end he arrives at the answer that
there are 42,100 different kinds of conditional. 12

A modern logician might make various distinctions and count in various
ways but, if he counted as Boethius does, he would have to allow that there are
a countably infinite number of kinds of conditional since there are no logical
limits to the complexity of propositional contents which may be the argu¬
ments of propostional functions.

Boethius tells us that his result would have been much larger if he had also
taken the quantity of the component categoricals into account. 13 He does not
do this he says because conditionals are generally stated indefinitely. For exam¬
ple, ‘ si est a, est b\ that is, ‘if something’s an a , then its a b\ or ‘if there’s an a ,
there’s b\ depending on whether or not the antecedent and consequent hold
of the same subject.

A conditional according to Boethius is thus distinguished from the corre¬
sponding categorical claim. Using an example from Cicero which will later
cause problems, he notes that when we say:

‘if someone is pregnant, then she has slept with a man we are not then saying that being-
pregnant is having-slept-with-a-man, but rather it is only proposed that pregnancy can
never occur without sleeping with a man. 14

n) DHS 1.6.5-1.7.5. Categorical propositions are either affirmative or negative. They may in
addition be modalised in some way. Boethius distinguishes three modes of necessity for each of
which there is the corresponding mode of possiblitly. (1) the necessity that a proposition has in
virtue of being true, ‘necesse est Sortes sedere, dum sedet’; (2) what we might call the necessity
of natural dependence, ‘necesse est hominem habere cordum est atque vivit’; (3) unconditional,
’universal and proper, necessity, ‘necesse est Deum esse immortalem’. He claims that possiblity
and necessity coincide in case (1) with actuality. This gives five modalities in all.

12) Note that Chrysippus too engaged in calculation but of different sort. He asked how many
distinct conditionals could be formed from a set of ten distinct simple propositions. This is a
combinatorial problem which when precisely formulated has a definite finite answer. See Stanley
(1997), 344-350, also available on line at http://mathworld.wolfram.com/PlutarchNumbers.
html.

,3) Note that there is no hint in Boethius that we might quantify the whole conditional.

14) DHS 1.2.1: ‘At in his propositionibus quae conditionales dicuntur non est idem praedicatio-
nis modus; neque enim omnino alterum de altero praedicatur, sed id tantum dicitur esse alterum,
si alterum fuerit, ueluti cum dicimus: “Si peperit, cum uiro concubuit”. Non enim tunc dicitur
ipsum peperisse id esse quod est cum uiro concumbere, sed id tantum proponitur quod partus
numquam esse potuisset nisi fuisset cum uiro concubitus.’

C. J. Martin / Vivarium 45 (2007) 153-168

15 7

Uttering an affirmative categorical we assert that there is something to which
the subject term applies and that the predicate term applies to it too. With the
corresponding conditional we indicate only that there is a connection between
their applicability without committing ourselves to their application. 15

According to Boethius the connection in question is the necessity of insep¬
arability. This relation between antecedent and consequent, he notes, is usu¬
ally indicated with ‘sf but sometimes ‘cum l(i may be used instead to form a
conditional. He claims, however, that although both of these conjunctions
indicate the inseparability of consequent from antecedent, there is a distinc¬
tion between them. With £ cum only accidental inseparability is indicated. The
heavens are necessarily spherical and fire is necessarily hot so ‘ cum ignis calidus
sit , coelum rotundum est' is a true hypothetical. It is not possible for either the
antecedent or the consequent to be false and so not possible for the antecedent
to be true and the consequent false.

Boethius maintains, on the other hand, that the conjunction l si 5 indicates in
addition to inseparability some kind of explanatory connection. In the case of
‘si est homo, est animal' , he tells us, the consequent explains the antecedent, since
it is the genus which is the cause of the species. In the case of £ si terrae fuerit
obiectus, defectio lunae consequitur the antecedent explains the consequent.

Unfortunately Boethius tells us no more about these explanatory connec¬
tions and seems here to suppose that inseparablity is both necessary and
sufficient for consequence. 17 Thus:

If someone wishes to oppose a conditional proposition he will bring it about that he destroys
the consequence. For example, he will not oppose the consequence ‘if a is, then b is’ if he

15) DHS 1.2.2: ‘In praedicatiua igitur id spectabimus quod ipse homo animal sit, id est nomen
in se suscipiat animalis, in conditionali uero illud intellegimus, quod si fuerit aliqua res quae
homo esse dicatur, necesse sit aliquam rem esse quae animal nuncupetur. Itaque praedicatiua
propositio rem quam subicit praedicatae rei suscipere nomen declarat; conditionalis uero propo-
sitionis haec sententia est, ut ita demum sit aliquid, si fuerit alterum, etiamsi neutrum alterius
nomen excipiat.’

16) DHS 1.3.1 ‘Possunt autem aliquando etiam hoc enuntiari modo: “cum hoc sit, illud est”
ueluti cum dicimus: “cum homo est, animal est” uel: “cum homo est, equus non est”; quae
enuntiatio propositionis eiusdem potestatis est cuius ea quae hoc modo proponitur: “si homo est,
animal est”, “si homo est, equus non est”.’

17) Matters are in fact a bit more complicated since consequence seems to include both insepa¬
rability and incompatibility; DHS 1.2.3: ‘.. .aut in conditione posita consequentia ui coniunc-
tionis uel disiunctionis ostenditur.’

158

C. J. Martin / Vivarium 45 (2007) 153-168

shows either that a does not hold or b does not hold. Rather he must show that if a is
posited b does not immediately follow, but that a may be even if b is not. 18

Boethius goes on, however, to claim that conditionals, like simple categoricals,
are either affirmative or negative and that the quality of a conditional is deter¬
mined by that of its consequent. Whatever the quality of its antecedent, a
simple conditional with an affirmative consequent is thus affirmative, one with
a negative consequent is negative, and likewise for composite conditionals.

Quality defined for conditionals in this way is, I think, an attempt to
make the account of simple conditionals accord with that of categorical
propositions. It also suggests that Boethius supposes that the negation of a
given conditional is the conditional whose antecedent is the same as but
whose consequent is the negation of the consequent of the original. This
suggestion is confirmed, as we will see by his account of the logic of com¬
pound conditionals. Despite this, however, it is clear from his repeated
appeal to examples that contradict it that Boethius does not accept the
principle of conditional excluded middle, the principle, that is, that an
affirmative and the corresponding negative conditional divide truth and
falsity between them. 19

It is Boethius’ apparent confusion of the criterion which he gives for the
quality of conditionals with his account of what is required to show that a
conditional is false which makes his logic so difficult for later writers.

Boethius spends the second and third books of DHS in setting out each of
the syllogisms which hold for the various kinds of indefinite conditional and
disjunction which he recognises. In every case the syllogism is stated generally,
for example ‘if it’s a , then it’s b , but it’s a\ therefore it’s b\ Although Boethius
says nothing about their application he seems to intended these hypothetical
syllogisms to be understood as schemata for inferences with respect to a given
subject, that is, for any x, if * is a , then x is b.

The one difficult aspect of Boethius’ account of simple conditionals and of
conditionals in general is his claim that a conditional (whether simple or com¬
posite) with a negative antecedent and an affirmative consequent is true if and

18) DHS I, ix, 6: ‘Si quis igitur recte conditional! propositioni repugnabit, id efficiet ut earum
destruat consequentiam, ueluti cum ita dicimus: “si a est ,b est”, non in eo pugnabit si monstret,
aut non esse a , aut non esse b , sed si posito quidem a , ostendit non statim consequi esse b, sed
posse cssca , etiamsi b terminus non sit.’

19) Just as he, like Aristotle, cannot in general accept that every categorical divides truth and
falsity with its negation. The principle fails in the case of indefinite categoricals—‘homo est
albus’ and ‘homo non est albus’ may both be true at the same time.

C. J. Martin / Vivarium 45 (2007) 153-168

159

only if its antecedent and consequent are immediately opposed to one another.
That is to say, if and only the terms are both exclusive and exhaustive. In that
case, in virtue of the nature of things (natura rerum ) rather than the condi¬
tional form of the proposition (complexio propositionuni ), in addition to modus
ponens and modus tollens , affirming the consequent and denying the anteced¬
ent are also valid argument forms. 20

In his discussion of mediate conditionals, however, Boethius argues, for
example that from ‘if its not a , then it’s b and if it’s b , then it’s c and ‘it’s not
a it follows that ‘it’s c but we cannot argue in this case by denying the ante¬
cedent since there are counterinstances. Let V, using Boethius’ example, be
‘irrational’, *£’ ‘rational’, and ‘c’ ‘animal’. We infer ‘if it’s not irrational,
then it’s an animal’ but the terms ‘irrational’ and ‘animal’ are not, according to
Boethius, related as immediate contraries. 21

Boethius allows only simple disjunctions in DHS and argues that a given
disjunction is equivalent to the conditional whose antecedent is the nega¬
tion of the first disjunct and whose consequent is the second disjunct. The
simple disjunction ‘aut est a aut est is thus equivalent to the conditional
‘si non est a est U and so appears as the major in four valid inferences—two
in virtue of the complexion and two in virtue of the terms. The three other
forms of disjunction each support only the two inferences in virtue of
complexion.

Boethius’ representations of all the various instances of the hypothetical
and disjunctive syllogism must, as I said, apparently be understood as sche¬
mata which are to be applied to given subject. It is not possible otherwise, I
think, to make sense of his requirement that a conditional appearing as a

20) Cf 2.3.7: ‘Cum enim dicimus, “si a non est, esse b ” si assumamus: “atqui est a nihil euenit
necessarium, ut uel sit b uel non sit, secundum ipsius complexionis naturam. Nam hie quoque,
ut in his in quibus in assumptione secundus terminus ponebatur, dicendum est secundum qui-
dem ipsius complexionis figuram nullum fieri syllogismum; secundum terminos uero in quibus
solis dici potest, necesse esse, si a fuerit, b non esse. In contrariis enim tantum, et in his imme-
diatis, id est medium non habentibus, haec sola propositio uere poterit praedicari, ueluti cum
dicimus: “si dies non est, nox est”; siue non fuerit dies, nox erit, siue nox non fuerit, dies erit, siue
dies fuerit, nox non erit, siue nox fuerit, dies non erit.’

21) Note in the text of DHS as we have it Boethius runs through all of the forms of the mediate
hypothetical and shows in each case, by giving an example, that affirming the consequent and
denying the consequent fail. At the end of the discussion of the mediate conditionals of the first
figure, however, he remarks that in six cases these argument forms hold in virtue of a property of
the terms. These passages are absent in one family of manuscripts, suggesting either that they were
suppressed—as Obertello supposes—or that they are a later addition. See DHS (430-431).

160

C. J. Martin / Vivarium 45 (2007) 153-168

component of a conditional must not itself be true. For example for a propo¬
sition of the form ‘si est a, cum sit b, est c to be a conditional according to
Boethius, being-tf must be a genuine condition for being-c being a conse¬
quence of being-£. In characterising these conditionals he often uses explana¬
tory adverbs and once speaks of the antecedent as the cause of the consequent. 22
He maintains that in order for such a proposition to be genuinely conditional
it must be possible for b to hold of something without c also holding of it and
so, by his criterion for the destruction of consequences, ‘ si est b, est c is false.
In order for an instance of modus ponens with a conditional of this kind as its
major to be sound, the premisses must thus be understood as applied to a
particular subject.

For example the syllogism ‘if Socrates is human, then if Socrates is animate,
Socrates is an animal, but Socrates is human; therefore if Socrates is animate,
then Socrates is an animal’. We have proved the inseparability in Socrates of
being animate from being an animal. The explanation is that he is human.

Note that Boethius’ worry here cannot be that a compound conditional
might satisfy the inseparability condition merely because of the truth-value of
one its components. If this were so he would require that a conditional ante¬
cedent is not false. He does not do this, however, but rather requires in the
case of both ‘if (if it’s a , then it’s b ), then it’s c and ‘if (if it’s a , then it’s b ),
then (if it’s c, then it’s d)' that the component conditionals are all false.

Boethius’ account of conditionals with simple conditionals as both ante¬
cedent and consequent is particularly curious. He gives as truth-conditions
for the first form that neither the antecedent ‘if it’s a , then it’s b' nor the
consequent ‘if it’s c , then it’s d f is true but adds that being c must follow from
being tf, and being d from being b. Thus ‘if (if it’s a human, then it’s a doc¬
tor), then (if it’s animate, then it’s an artificer)’ is true, but the conditionalisa-
tion of the contraposition of a simple conditional, ‘if (if it’s a, then it’s b ),
then (if it’s not b , then it’s not a)\ which Boethius elsewhere characterises as
a consequence, is not.

Boethius describes the operation of modus tollens for simple conditionals in
terms of the destruction of the antecedent following from the destruction of
the consequent and continues to employ this terminology and that of ‘negat-

22) DHS , 3.9.1: ‘Nam, ut superius dictum est, non sufficit quolibet modo iungere terminos, ut
fiant hypotheticae propositiones ex duabus conditionalibus coniugatae; neque enim si quis dicat:
“Si cum homo est, animal est, cum dies est, lucet” talem fecerit propositionem quae ex duabus
conditionalibus constet, idcirco quia prior conditio non est secundae causa conditionis. Hoc
igitur superius positarum propositionum ratio demonstrat, quemadmodum fit ut conditionem
conditio consequatur. 5

C. J. Martin / Vivarium 45 (2007) 153-168

161

mg when giving his account of compound conditionals. 23 The puzzle is that
he treats the destruction of an affirmative conditional as equivalent to the
assertion of the corresponding negative and vice versa.

There is no problem, of course, as long as the antecedent is contingent,
with arguing from the truth of a negative conditional to the falsity of the cor¬
responding affirmative since they are contrary. In his account of the syllogism,
however, Boethius treats negative and affirmative conditionals as if they were
contradictory. For example in the schema, ‘if (if its a , then its b ), then (if its
c, then it’s d ), but (if its c, then its not d)\ therefore (if its a, then its not by.
On his own account of the destruction of a conditional, however, what
Boethius should conclude is only that its being a is not inseparable from being
b for the subject in question.

This, then, was the account of the logic of the conditional which was passed
down from antiquity to the philosophers of the twelfth century. Let us now
consider how Abaelard responded to it.

Abaelard

In the long treatment of the topics in his Dialectica Abaelard develops a sophis¬
ticated account of conditional propositions. In particular he takes the satisfac¬
tion of Boethius’ inseparability requirement to be necessary but not sufficient
for the truth of a conditional, and notes in passing that the condition is
satisifed by any conditional with an impossible antecedent. He does not
explicitly state the principle that anything follows from an impossibility, but
he could hardly be expected to do so since he does not think that following is
merely inseparability.

In order for a conditional to be true, Abaelard insists, there must be in addi¬
tion to inseparability a genuine connection between the antecedent and con¬
sequent. The required connection exists, he thinks, when the meaning of the
consequent is contained in that of the antecedent.

23) For example in discussing the fifth form of the second kind of composite conditional: DHS,
III, iv, 3: ‘Ex quinta etiam propositione ita syllogismi fiunt: si non est a, cum sit b, est c; atqui
non est a; cum igitur sit b, est c; uel ita: atqui est a, cum igitur sit b, non est c; uel ita: atqui cum
sit b, non est c; est igitur a; uel sic: atqui cum sit b, est c, non est igitur a. Quod idcirco euenit ut
huiusmodi propositio quatuor colligat syllogismos, quia in his tantum si non sit aliquid esse
aliud proponi potest, in quibus contraria medietatibus carent; in his enim uel interempto altero
alterum ponitur, uel posito altero alterum necesse est perimaturl Compare DHS III, vi, 3: ‘In
omnibus enim si quidem uelimus astruere, primam totius propositionis assumemus partem, si
uero in conclusione aliquid destruendum est, secunda negabitur.’

162

C. ]. Martin I Vivarium 45 (2007) 153-168

Abaelard believes that his account of the semantics of the conditional gen¬
erates what we now call a connexive logic, a logic, that is, for which no propo¬
sition can entail or be entailed by its contradictory opposite. He concludes
that no simple conditional whose antecedent and consequent are of a different
quality can be true and so that neither the relation of immediate contrariety
nor any other kind of opposition can warrant the truth of a conditional. It
follows that a hypothetical syllogism with such a conditional as its major
premiss will be valid but never sound.

Abaelard also notes that it follows from his definition of consequence in
terms of containment of sense that only one kind of affirmative categorical can
entail a conditional:

We concede no categoricals to be antecedent to hypotheticals other than those which show
the nature of things, which preserve for ever the necessity of consecution as, for example,
‘animal is the genus of human’ is <antecedent> to the consequence ‘if something is
human, then its an animal’. 24

Since conditionals are true, according to Abaelard, whether or not there exist
things of which they might be asserted, the only categoricals which can follow
from conditionals are those negative propositions whose truth is compatible
with the non-existence of their subjects. His example is that the categorical
no human is a stone follows from the conditional ‘if something is human,
then its not a stone’. Within Abaelards connexive logic this latter conditional
is false, and necessarily so, but if it were true, then it would follow in virtue of
the meaning of the conditional that no human is a stone.

Abaelard goes on to explicitly contrast his own view with that of the major¬
ity who he says hold that conditionals are equipollent to the corresponding
universal categoricals. 25

Conditionals whose antecedent and consequent are both simple condition¬
als are much happier constructions for Abaelard since he believes that every

24) Abaelard, Dialectica , III. 1 (Peter Abelard [1970], 283 [all page references are to this edition]):
‘Unde non alias categoricas ad hypotheticas antecedere concedimus, nisi quod rerum naturam
ostendant, quae consecutionis necessitatem in perpetuum cus<to>diant, ut istam: “animal est
genus hominis” ad huiusmodi consequentiam: “si est homo, est animal”.’

25) Dialectica , IV. 1 (475-476): ‘Affinitas autem tanta est istarum <cum> contrariis ut fere ab
omnibus mutuam ad eas inferentiam habere concedantur. Has enim inuicem aequipollere secun¬
dum mutuam inferentiam annuunt. Hinc quidem “omnis homo est animal” ac “si est homo, est
animal” illinc uero “nullus homo est animal” ac “si est homo, non est animal”. Nos tamen qui,
ut in sequentibus apparebit, has omnes inferentias non recipimus, aequipollentiam uerae infer-
entiae in eis non concedimus.’

C. J. Martin / Vivarium 45 (2007) 153-168

163

simple conditional entails its contrapositive. 26 Unfortunately, as we saw,
Boethius will allow neither Abaelard’s compounds of conditional and categor¬
ical nor his conditionals with both antecedent and consequent conditional
as the major premisses of a hypothetical syllogism since he requires that the
component conditionals in such propositions are not themselves true. For
Abaelard, furthermore, the truth of a conditional which is warranted by a
relationship between predicate terms is independent of the subject of which it
is asserted. If he followed Boethius, valid arguments which detach conditional
consequents from their antecedents could not possibly be sound. 27

Abaelard is thus presented with an account of the logic of the conditional a
good part of which can be of no use to him beyond providing a purely formal
exercise. And even here, I think, it is ultimately of no interest. Abaelard clearly
has concepts of propositional content and propositional operation like our
own, and thus formulates the primitive inference rule modus ponens and the
derived rule modus tollens in a completely general way, applicable to condi¬
tionals of any degree of complexity. 28

Presented with Boethius’ remarks on the destruction of consequences and
negative conditionals Abaelard appeals to a distinction which he first makes
for categorical propositions between negation as a propositional operation,
what he calls destructive negation, and separative negation.

The destructive negation of^a given proposition is simply the proposition
which is true if and only if the proposition negated is false. It is marked by
preposing the negative particle to the proposition to be negated, and it is iter-
able without limit. Separative negation, on the other hand, is defined on a case
by case basis for different kinds of propositions. Abaelard follows Boethius in
characterising a conditional as affirmative if and only if it has an affirmative
consequent. 29 Only affirmative conditionals have separative negations and

26) I.e. (p->q) <-> (not:q->not:p).

27) Suppose Abaelard accepted that l a->(b-c)’ is true only if ‘b->c is false. ‘b->c’ then cannot be
true of any subject. But then although the inference a->(b->c), a |- (b->c)’ is valid can never be
sound.

28) Dialectica , IV, I (500-501).

29) See Dialectica , II, ii (176-177): ‘Multum enim refert ad sententiam enuntiationis, cum prae-
posita negatiua particula totam exigit et destruit afhrmationem, et cum eadem interposita termi-
norum separationem facit, quod quidem ex hypotheticis quoque enuntiationibus ostenditur.
Non enim eadem est sententia istarum: (1) “si est homo, non est iustus” et (2) “non si est homo,
est iustus”. Ilia namque demonstrat hominis positionem non pati iustum, haec uero non neces-
sario exigere iustum; quod uerum est, illud autem falsum. Et haec quidem quae negatione prae-
missa totam hypotheticam perimit, hanc scilicet: (3) “si est homo, est iustus”, eius propria
negatio dicitur ac recte diuidens, quae scilicet nec uera simul cum ea nec falsa esse potest, quippe

164

C. J. Martin / Vivarium 45 (2007) 153-168

these are obtained by negating the consequent. 30 Where it makes a difference,
the negation of the consequent may be either its destructive or separative
negation. 31 The second is more opposed to the affirmative conditional than
the first and properly speaking its contrary.

In order to make something of the inferences set out in DHS\ Abaelard
draws attention to Boethius’ method of constructing compound conditionals.
The principle connective here is always ‘si and the embedded conditional is
always formed with ‘cum . Since for consequence Boethius apparently requires
only inseparability, this is probably no more than a stylistic point but Abaelard
connects it to the distinction between explanatory and non-explanatory con¬
sequences to make a contrast between conditionals and what he calls temporal
propositions. According to Abaelard, translating ‘curri as ‘when:

Boethius called those propositions ‘temporal’ in which something is conjoined to some¬
thing temporally and not conditionally, as for example: when it rains, there is thunder’.
For here it is not proposed that ‘if it rains, then there is thunder’, but rather ‘when it rains
it also thunders’, that is, that at the time at which one occurs (<quo tempore ), so also does the
other, as if they were said to occur simultaneously. 32

Abaelard goes on to say that temporals hold so long as their components hold
together, independently of whether they are able to occur apart, or whether one

eius sensum simpliciter destruit. Ilia uero simul esse falsa potest, numquam autem simul uera.
Unde potius contraria ei uidetur quam contradictoria. Sic quoque in categoricis propositionibus
ea tantum propria contradictio ac recte diuidens cuilibet affirmationi uidetur quae negatione
praeposita totam eius sententiam destruit, ut eius quae est: “omnis homo est homo” ea quae est:
“non omnis homo est homo”, non ea quae est: “quidam homo non est homo”; haec enim fortasse
simul erit falsa cum ea. Re enim hominis prorsus non existente neque ea uera est quae ait: “omnis
homo est homo” nec ea quae proponit: “quidam homo non est homo”, hoc est “quaedam res
quae est animal rationale mortale, non est animal rationale mortale”; “hominis” enim nomen
nonnisi ex praesentia animalis rationalitate et mortalitate informati impositum fuit.’

30) Dialectica , IV, i (478): ‘Cuilibet enim propositioni ad totam eius sententiam perimendam
negatio praeponi poterit; aeque quidem et ad “Socrates est homo” et ad “Socrates non est homo”.
Omnis itaque propositio negatiuam destructoriam habere poterit, sed non omnis separatiuam
nisi sola affirmatiua: terminorum enim est separatio, totius uero propositionis destructio.’

31) The separative negation of a simple proposition entails its destructive negation so Abaelard
argues that the ‘separative negative’ for a given affirmative is more more opposed to it than the
destructive negative.

32) Dialectica , IV, i (473): ‘Temporales uero illas Boethius uocauit quae temporaliter, non condi-
tionaliter, aliquid alicui coniungunt, ut ista: “cum pluit, tonat”. Non enim hoc loco proponitur
quod si pluit, tonat, sed magis id dicitur quod quando pluit, et tonat, id est: “quo tempore unum
contingit, et alterum”, ac si uidelicet utrumque simul fieri diceretur.’

C. J. Martin / Vivarium 45 (2007) 153-168

165

requires the other. We cannot immediately conclude, however, that there is no
necessity here since he also characterises the truth of temporal propositions as
requiring concomitance rather than consecution; consecution and concom¬
itance’, though without qualification, are the terms he uses in contrasting the
necessity of entailment with the weaker necessity of mere inseparability.

Furthermore, in the passage just quoted at the time’ translates ‘quo tem¬
pore which is precisely the phrase which Boethius uses to characterise the
inseparability of the heavens being spherical from the hotness of fire, 33 Abae-
lard’s observation that the components of the temporal might hold apart from
one another might thus be a reference to merely conceptual possibility. 34

If this were so, however, Abaelard would be committed to identifying
omnitemporal coincidence with inseparability and possibility with actuality at
a time. I have argued elsewhere, however, that there is strong evidence that he
does not understand possibility in this way. 35

That he does not do so here is confirmed by his later explicit statement of
the truth conditions for temporal propositions. The compound temporal
proposition formed from the two propositions ‘Socrates was a youth’ and
‘Socrates was an infant’ is true at a given time, he tells us, just in case each
component is true at that time, and so true now of the old Socrates. The
proposition is not true at every time, however, and for some times one com¬
ponent is true and the other false.

Abaelard thus does not commit himself with his example to the omnitem¬
poral coincidence of rain and thunder and the identification of possibility
with actuality at a time. Since he holds, however, that what is true of the pres¬
ent is necessarily true of it, a true present tense temporal will in that sense be
necessary. All the material needed for the ut nunc conditional is here though
Abaelard himself does not use the expression.

33) DHS, I, iii, 6: ‘Secundum accidens hoc modo, ut cum dicimus: “cum ignis calidus sit, coelum
rotundum est”. Non enim quia ignis calidus est, coelum rotundum est, sed id haec propositio
designat, quia quo tempore ignis calidus est, eodem tempore coelum quoque rotundum est.’

34) Abaelard s next remark seems to tell very much in favour of a reading which distinguishes
between temporal coincidence and necessary comitatio without consequence: Dialectica , IV, i,
pp. 482-483: ‘In his autem quarum consecutio nihil aliud est dicenda quam in eodem tempore
comitatio, nulla est consecutionis natura pensanda; sed dum membra sint uera, et ipsam esse
ueram consequentiam concedunt, alioquin falsam; et indifferenter alterum ad alterum et antece-
dere potest et consequi.’ ‘ Concedunt ’ seems wrong here since there is no reference anywhere to
those who might be considering this kind of proposition. Perhaps it should be ‘concedenda .

35) Martin (2004).

166

C. J. Martin / Vivarium 45 (2007) 153-168

Abaelard notes that temporal propositions formed with ‘cum may be con¬
strued either as simple categoricals or as compound propositions. The first
reading is the one which interests him since with it he can at least appear to
agree with Boethius on modus tollens for compound hypothetical.

The question of how to understand the negation of a conditional had obvi¬
ously become a pressing one when Abaelard wrote. Faced with Boethius’ claim
that the antecedent of a compound conditional must be a genuine condition
for the truth of the consequent and his use of the separative rather than the
destructive negation, some logicians, Abaelard tells us, had construed the
propositions embedded in compound conditionals as temporal compound
propositions. In such conditionals they understood the consequential relation
to be between the categorical component and the consequent of the embed¬
ded temporal compound.

As when it is said ‘if (when its animate its human), then it’s an animal’ the force of entail-
ment is between human and animal, and animate plays no role in the entailment but is set
down as a certain accompaniment, just as in the case of the consequence ‘if it’s human and
a stone, then it’s an animal’, where the entire force of entailment is between human and
animal. 36

On this interpretation the antecedent of the embeded proposition is, as it
were, held fixed in modus tollens. Since for the truth of‘if (when its a its b ),
its c being b must entail being r, if c is not predicated of a given subject,
they supposed, we may infer that neither is b but that nevertheless the sub¬
ject remains an a.

Abaelard notes, however, that to save Boethius those who interpreted the
inference in this way were forced to substitute ‘quamvis for ‘cum and to
draw the conclusion that although its a , its not b? 7 Their interpretation fails
he argues because the substitution of ‘quamvis for ‘cum will often result in a
conclusion which is necessarily false. If, for instance, we take stone as the sub¬
ject of the above inference we may argue: if (when a stone is animate, it is
human), then it is an animal, but a stone is not an animal; therefore although

36) Dialectica , IV, i (484): ‘Veluti cum dicitur: “si (cum est animatum est homo), est animal” inter
“hominem” et “animal” uis inferentiae consideratur et “animatum” quidem ad inferentiam non
operatur, sed ad quamdam comitationem ponitur, ueluti si tabs fieret consequentia: “si est homo
et lapis, est animaTcum inter “hominem” solum et “animal” tota uis inferentiae penderet.’

37) Dialectica , IV, i (484): ‘Ideoque in destructione “cum” in “quamuis” conuertunt, ac si ita dice-
rent: “quamuis etiam esset animatum”. Alioquin Boethium non possunt absoluere qui tantum ad
sequentem partem temporalis negationem apponit, secundum id scilicet quod ad earn uis infer¬
entiae respicit, siue scilicet temporalis ad categoricam siue categorica antecedat ad temporalem.’

C. J. Martin / Vivarium 45 (2007) 153-168

167

a stone is animate, it is not human. From this conclusion we may infer that a
stone is an animal.

Furthermore, Abaelard notes, 38 the proponents of this theory are commit¬
ted to an account of modus tollens for conditionals with both antecedent and
consequent conditional which is incompatible with that given by Boethius. If
what is important in deciding the form of the conclusion are the terms between
which the relation warranting the entailment holds, then ‘if (when its a it’s b ),
then (when it’s c it’s d ), but when its not c it’s d\ therefore when it’s not a it’s
b\ is as much an instance of modus tollens as the form given by Boethius since
for the truth of the conditional a must entail r, and b must entail d ’

The fundamental problem with this theory, however, is that it fails to locate
the relation of entailment at stake in modus tollens in the proper place. What
is asserted in asserting a conditional, as Abaelard with his clear understanding
of propositionality explicitly recognises, is that the consequent conditional
follows from its antecedent. If the antecedent is compound then the conse¬
quent follows from the whole compound even though the conditional may be
true because of a relationship between only one component of the antecedent
and the consequent of the whole conditional. 39

The only way to solve the problems posed by Boethius’ text, according to
Abaelard, is to emend. To construe the temporal components of putatively
compound conditionals as categoricals and so the conditionals themselves a
simple. Using ‘quando (‘when’) rather than ‘cum he thus proposes that we
analyse the temporal proposition ‘fire is hot when the heavens are spherical’
as having ‘fire’ for its subject, ‘is hot’ for its predicate and ‘when the heavens
are spherical’ as a determination of the predicate, thus:

[ignis] subject —[quando caelum est rotundum] Determination —(est)[calidus] Predicate ] 40

38) Dialectica , IV, i (487): ‘Amplius: si secundum uim inferentiae destructiones in his oporteat
pensare, quare in his hypotheticis quae ex duabus temporalibus iunguntur, in quibus duplex uis
inferentiae consideratur, non sint duplices destructiones? Veluti cum ita proponitur: si cum est
homo est medicus, cum est animal est artifex ‘homo’ enim ad ‘animal’ et ‘medicus’ ad ‘artificem’
tamquam ad sua tota referuntur.’

39) Dialectica , IV, i (485-486): ‘Primumque illam calumniemur defensionem quam praetendunt
de destructione secundum principalem inferentiam, ueluti cum dicitur: “si (cum est animatum
est homo), est animal” hie enim uim inferentiae inter “hominem” et “animal” considerant; nihil
enim magis ‘animatum’ quam quislibet alius terminus operatur. Sed dico quod licet uis inferen¬
tiae in terminis consistat, tota tamen est propositio destruenda.’

40) Abaelard employs this analysis of categoricals elsewhere in his analyses of modal propositions
and of ‘qualifications’ ( constantia ) in explaining how the to qualify conditionals connecting
immediate opposites to avoid paradoxes.

168

C. J. Martin I Vivarium 45 (2007) 153-168

In negating this categorical, Abaelard insists, we must remove both the
predicate and the determination. In the problem case of ‘if (when a stone is
animate, its human), then it’s an animal’, the negation of the antecedent
understood as a temporal categorical is thus the claim that a stone does not
have the property of being human-when-it-is-animate, which does not entail
that it has the property of being animate.

If we treat temporals as categoricals in this way and read each of Boethius’
embedded hypothetical as a temporal categorical, then his various strange
hypothetical syllogisms are valid. They are valid, however, only because they
are no longer strange but simply so many substitution instances of modus
ponens and modus tollens with the conditional connective and negation under¬
stood as propositional operations.

Abaelard thus has a cure for Boethius but in the end he does not apply it.
Rather he adopts the alternative emendation which does not even save the
appearance of Boethius’ arguments. In glossing the all the various kinds of
hypothetical syllogism provided by his ancient authority Abaelard reads the
embedded hypotheticals as compound temporal propositions and in every
case replaces the negation of the consequent with the negation of the whole
compound.

Again Boethius’ syllogisms become valid arguments but again they are no
more than substitution instances of modus ponens and modus tollens con¬
structed by a logician who has a clear understanding of propositional form.

If Abaelard could not save Boethius, then no one can. The logic of the con¬
ditional is straightforward enough when one properly understands it as a
propositional connective. The basic inference forms are, modus ponens and
transitivity , with modus tollens as a derived rule. There is no special place in
such a logic for Boethius’ curious conditionals and no place at all for his con¬
sequent negation. It is not surprising, then, that a century after Abaelard had
demonstrated this there was no place for Boethius’ book on the hypothetical
syllogism in list of required reading at Paris.

BRILL

Vivarium 45 (2007) 169-188

VIVA

RIUM

www.brill.nl/viv

Are Thoughts and Sentences Compositional?
A Controversy between Abelard and a Pupil
of Alberic on the Reconciliation of Ancient
Theses on Mind and Language*

Martin Lenz

Humboldt-Universitat zu Berlin

Abstract

This paper reconstructs a controversy between a pupil of Alberic of Paris and Peter
Abelard which illustrates two competing ways of reconciling different ancient tradi¬
tions. I shall argue that their accounts of the relation between sentences and thoughts
are incompatible with one another, although they rely on the same set of sources. The
key to understanding their different views on assertive and non-assertive sentences lies
in their disparate views about the structure of thoughts: whereas Abelard takes thoughts
to be compositional, the opponents arguments seem to rely on the premise that the
mental states which correspond to sentences cannot be compositional in the way that
Abelard suggested. Although, at a first glance, Abelards position appears to be more
coherent, it turns out that his opponent convincingly argues against weaknesses in
Abelard’s semantic theory by proposing a pragmatic approach.

Keywords

compositionality, semantics, language, assertive and non-assertive language, thought

Introduction

A weak but general notion of compositionality can be summarised as the idea
that the meaning of a sentence is determined by the meaning of its parts.

+) I am very grateful to Margaret Cameron, John Marenbon and Irene Rosier-Catach for dis¬
cussing my views and for letting me have copies of their papers from 2004 in advance of publica¬
tion. I would also like to thank the Deutsche Forschungsgemeinschaft for the generous support
of my studies.

DOI: 10.1163/136833407X217704

170

M. Lenz i Vivarium 45 (2007) 169-188

Although medieval philosophers did not endorse the stronger tenet (held by
Frege) that words have meaning only in the context of a sentence, most medi¬
eval philosophers held some version of the weaker claim, which is implicit in
the logic of Aristotle . 1 If, however, one wants to explicate in what way sen¬
tences and the corresponding thoughts are compositional, one faces enormous
difficulties, many of which are still debated. Can thoughts and sentences be
structured in similar ways? Can we ascribe semantically complex structures to
ontologically simple units? Do thoughts that correspond to assertive sentences
share the contents of thoughts that correspond to non-assertive sentences?

In the twelfth century there is an increasing number of conflicting commen¬
taries on Aristotles De interpretatione , and, as is well known, especially Peter
Abelard s commentary was intensely challenged for various reasons. Like Aris¬
totle himself, most medieval logicians of this time and later periods discussed
the central topic of De interpretatione , i.e. the notion of the sentence, with
regard to truth and falsity, thereby highlighting the assertive sentence or state¬
ment (proposition oratio enuntiativa ) in their semantic analyses. But in focusing
on the question of how to give an adequate definition of the assertive sentence,
the problem emerged how to distinguish the assertive sentence from other
types of sentences (namely non-assertive sentences). Aristotle had tackled, or
rather avoided, this problem by simply excluding non-assertive sentences from
the realm of logic . 2 By contrast, the Stoics defended a view which explicitly
includes the lekta of non-assertive sentences in their analyses, and although
Abelard and his contemporaries knew very little of Stoic philosophy, parts of
its positions had been transmitted through the well known writings of author¬
ities such as Augustine, Boethius and Priscian . 3 The early medieval writers were
well aware of the problems posed by Aristotles exclusion, and tried to fill the
gaps by reconciling Aristotles theses with the logical and grammatical tradi¬
tions that account for the relation between the different types of sentences.

In this paper I would like to introduce a controversy between a pupil of
Alberic of Paris and Abelard which shows two very disparate ways of reconcil¬
ing the different traditions. I shall argue that their accounts of the relation
between sentences and thoughts are incompatible with one another, although
they rely on the same set of sources. The key to understanding their different
views and to seeing how they combine the ancient views is, I believe, the fact

11 See Meier-Oeser (2004), 321-322.

2 > See Aristotle (1966), 17al-7.

’» See e.g. Long and Sedley (1987), 199-201.

M. Lenz / Vivarium 45 (2007) 169-188

171

that they have disparate views about the structure of thoughts ( intellect ™):
whereas Abelard takes thoughts or understandings of sentences to be compo¬
sitional, the opponent’s arguments seem to rely on the premise that the mental
states which correspond to sentences are not compositional, or at least not
compositional in the way that Abelard suggested. The question of the compo-
sitionality of mental states is not explicitly answered in Aristotle’s De interpre-
tatione. And one might read the theses of later authors such as Boethius and
Priscian, who invoke Stoic ideas on the subject, as an attempt to settle this
question.

To introduce the problem under discussion I would like you to consider the
following situation. Suppose you happen to bump into an old friend in the
street. The very moment he spots you he starts trembling all over. What would
you think? Imagine that you have lent him money some time ago, then you
might think something along the lines of “He is obviously scared”—“Scared
of what?”—“Maybe he is afraid that I might want my money back.” Your
friend might say to you: “If only I had some money on me!” Or: “I fear that
I haven’t got enough money on me.” Or: “If I had enough money on me,
I would give it back to you right away.”

Some models of language and mind seem to offer quite coherent accounts
of such a situation. One that might come to mind immediately is the idea of
ascribing propositional attitudes in relation to propositional contents. Accord¬
ingly, one could ascribe the attitude of desire or fear to a person in relation to
the content “that he hasn’t got enough money.” If one intends to ascribe such
an attitude to someone, one needs some sort of unit to relate it to—ideally a
unit of language. But the talk of units is prejudiced: we must have an opinion
about what constitutes a unit and what doesn’t, and as soon as there are parts
involved and we are not solely relying on (ontologically) simple units, we must
have some notion of a complete unit.

In contemporary philosophy of language and mind the notion of perfection
or completeness plays an important role with regard to determining speech
acts, thoughts, propositions (understood as significates of sentences) and such
like. The central idea of speech act theory, usually traced back to J. Austin and
J. Searle, is that sentences can be construed as expressions of speech acts and
analysed into two components: (1) the illocutionary act and the (2) proposi¬
tional content of the illocutionary act. Accordingly, the same propositional
content can be expressed with an assertive mode or a non-assertive mode, so
that all sentences can be said to have the same structure, consisting of the
(1) mode-indicator (or illocutionary force indicator) and the (2) propositional
indicator. When Searle developed this theory and even when he several years

172

M. Lenz / Vivarium 45 (2007) 169-188

later put the distinction between illocution and content to use with regard
to the mental sphere—distinguishing between (1) psychological mode (prop¬
ositional attitude) and (2) intentional content—, the notion of completeness
must have been quite uncontroversial, for he writes without further explana¬
tion: .. the grammatical clothing of an illocutionary act is the complete

sentenced

Any history of medieval philosophy with so much hindsight is of course
dangerously prone to anachronism. But precisely for that reason it might be
helpful to give some direction in the way of marking the starting point. As I
have shown elsewhere, the twelfth century philosopher Peter Abelard defended
a theory that resembles modern accounts of propositional attitudes and that
goes far beyond the established Aristotelian conceptions of the sentence. 4 5 It is
of course possible to challenge such a way of analysing sentences on various
grounds. Is it, for instance, plausible to assume the suggested correspondence
between language and thought? In the example given above it is suggested that
both the trembling and the exclamation (“If only I had...”) express a mental
state in relation to a content. But can we really assume that the speaker (your
friend) thinks the same as the hearer (you), or that he has cognitive access to the
content that his trembling and his utterance of the sentence is supposed to
express? Can we ascribe contents to a trembling? One might claim that at least
the sentences “If only I had some money on me!” “I fear that I haven’t got
enough money on me” and “I haven’t got enough money on me” are compo¬
sitional in the sense that they are related to contents which could be called
identical in a certain way. But can we equally claim that the thoughts of the
speaker are the same, let alone the truth conditions? And do the speaker and
the hearer have the same thoughts? Do they have access to the same con¬
tents?

In what follows, I would like to confront Abelard’s theory with the critique
of an Albricanus (i.e. an anonymous pupil of Abelard’s opponent Alberic of

4) Searle (1969), 25 (italics mine); cf. ibid. 23 (italics mine): ‘Thus we shall say that in the utter¬
ance of all four [sc. ‘Sam smokes habitually. 5 , ‘Does Sam smoke habitually?’, ‘Sam, smoke habit¬
ually! 5 , ‘Would that Sam smoked habitually. 5 ] the reference and predication are the same, though
in each case the same reference and predication occur as part of a complete speech act which is
different from any of the other three.’

5) Cf. Lenz (2005). I hasten to add that—although Abelard distinguishes between propositional
attitudes {affectus animi) and contents ( intellectus) —he does not develop an exhaustive theory of
speech acts.

M. Lenz / Vivarium 45 (2007) 169-188

173

Paris) in the light of the different traditional theses that both authors were
drawing on. I shall then argue that Abelards semantics of sentences is coun¬
tered with a communicative account of sentences, which seems to block Abe¬
lard’s idea of compositionality. Finally, I would like to draw some tentative
conclusions about the outcome of the controversy between Abelard and the
Albricani.

Traditional Sources of the Debate

It is plausible to assume that the anonymous author studied logic in Paris
around 1130. His Glossae “doctrinae sermonum” obviously reflect the teachings
of Abelard and Alberic of Paris, but show a strong inclination towards Alberic’s
opinions, which is why we might refer to him as one of the Albricani. The
controversy between Alberics anonymous pupil and Abelard was first noted
by Lambertus de Rijk, 6 and it has been brought to our close attention through
a very influential article by Klaus Jacobi, Christian Strub and Peter King.
However, it has never been studied in its own right. 7

A comparative reading of Abelard’s Logica and the Albricanus* Glossae “doc-
trinae sermonum” immediately reveals two striking facts: (1) the anonymous
author is very familiar with Abelard’s text, and (2), although he discusses the
very issues that “would prompt Abelard to develop” 8 the famous notion of the
dictum , 9 he does not mention the term once. What are we to make of this
omission? A well supported conclusion is that the Albricanus simply did not
know the term, because, as John Marenbon has convincingly argued, at this
late stage Abelard might have re-used the teaching-material from his Logica
without referring to dicta anymore. 10 This conclusion, however, does not rule
out another possibility which arises if we take the anonymous author’s seman¬
tic approach seriously: the Albricanus might have had no use for the notion
within his semantic framework, no matter whether he knew the term or not.

6) See De Rijk (1966); cf.. 46ff for transcriptions from the Glossae ‘doctrinae sermonum. A
critical edition of the text is currently being prepared by Peter King.

7) See Jacobi, Strub and King (1996), 15-40. Their theses are challenged in Rosier-Catach
(2004) and Marenbon (2004a).

8) Jacobi et al. (1996), 37.

9) The dictum -theory has received very much attention; a concise survey of various interpreta¬
tions is provided by Marenbon (1997a), 202-209. In short: the dictum is that which is stated by
an assertive sentence.

10) Cf. Marenbon (2004a).

174

M. Lenz / Vivarium 45 (2007) 169-188

Now, this conclusion gains some support if we assume that the controversy is
not about dicta at all, but rather about Abelard’s distinction between affective
propositional attitudes (affectus) and their contents ( intellects ), and thus
about the question whether the understandings or thoughts of sentences are
compositional in the way that Abelard suggested.

To endorse this interpretation, I would like to turn briefly to the back¬
ground of the controversy, so as to name the “main ingredients” of Abelard’s
and the Albricanuss positions. An obvious source is Aristotle, since both works
are commentaries on his De interpretatione . But both authors also refer to
theories, or rather parts of theories which come from quite different logical
and grammatical traditions, such as the ancient semantics of verbs, Priscian’s
definition of the sentence, and Boethius’s version of the division between
complete and incomplete sentences. 11

The grammarian Priscian—strongly influenced by the Stoics—adopted the
position of Apollonius Dyscolos according to which the different modes of
verbs express different emotional or volitional inclinations of the speaker’s soul
(Modi sunt inclinationes animi y varios eius affectus indicantes) } 2 That means
that if I utter a sentence such as “I haven’t got enough money” using the
indicative mode, I simply make an assertion, but if I utter the sentence “If
only I had enough money!” I express an attitude, namely a desire. Another
ingredient is Priscian’s definition of oratio, according to which a sentence is a
congruous order of words, demonstrating a complete sense’ ( oratio est ordina-
tio dictionum congrua , perfectam sententiam demonstrari)P A third ingredient
is Boethius’s division of complete and incomplete sentences, according to
which an incomplete sentence such as “a running man” leaves the hearer in
suspense, whereas, on hearing a complete sentence such as “a man runs”, the
hearer’s mind rests. 14

The central distinction that Abelard employed in his account of proposi¬
tional attitudes and their contents is the division between complete and incom¬
plete sentences ( orationes perfectae et imperfectae) . Although there are different

n) See Ebbesen (1982), 101-127, cf. esp. 109-111.

12) Priscian (1855), 421. The interpretations of this tenet in the tradition of the grammatica
speculativa are quite different from the interpretations discussed here: see Kelly (2002) 124-127.

13) Priscian (1855), 53. Cf. Kelly (2002), 154ff.

14) See Boethius (1880), 8-9: ‘Orationum uero aliae sunt perfectae, aliae imperfectae. Perfectae
sunt ex quibus plene id quod dicitur ualet intellegi; imperfectae in quibus aliquid adhuc plenius
animus exspectat audire...’

M. Lenz / Vivarium 45 (2007) 169-188

175

criteria for distinguishing those, it is—at least by the time of Abelard—
generally agreed that a complete sentence contains a finite verb (e.g. homo cur-
rit ), whereas an incomplete sentence does not (e.g. homo currens.y 5 Now, where
did Abelard adapt this distinction from?

It is rightly claimed that the division as such dates back at least to the Peri¬
patetic and Stoic philosophers and was taken up and transmitted by authors
such as Augustine, Boethius, Ammonius and Al-Farabi. 16 As could easily be
demonstrated, Aristotle’s and Priscian’s models for distinguishing the different
units of language are not as compatible as some medieval logicians and gram¬
marians might have claimed or wished, since they rely on different criteria,
such as truth and falsity (Aristotle) and completeness (Priscian).

In this respect it is interesting to note that the Stoic notion of the lekton is
quite often, if tentatively, compared to the notion of the dictum , the significatum
propositions, the enuntiabile or the complexe significabile. But already at this
point it should be clear that—even if one puts the question of evidence for
direct historical links to the Stoics aside—the comparison falls short, since the
lekton covers more types of sentences than the dictum (which only relates to
what statements say). The Stoics construed the notion of lekton not only with
regard to true or false statements but also with regard to non-assertive sen¬
tences such as wishes and commands. 17 Augustine cites the distinction between
complete assertive and complete non-assertive sentences (in the very text in
which he refers to the Stoic lekton by introducing the term dicibile) in his De
dialectics 18

But Abelard quotes neither the Stoics nor Augustine explicitly in this con¬
text, even if we can be certain that he adopted many Augustinian ideas. In his
discussions of the notion of sentence ( oratio ), both in the Dialectica and the
Logica Ingredientibus , he mainly refers to Aristotle, Boethius and Priscian.
Their positions are well known and extensively discussed in the early Middle
Ages, particularly in the various redactions of the Glosule super Priscianum
which circulated in the early twelfth century, and some of which are linked to
Abelard’s teacher and opponent William of Champeaux. 19

15) Some previous opinions reported in the Glosule super Priscianum from early tweltfth century
still disagree about this issue. Cf. Rosier-Catach (2004).

16) See Schneider (1993); cf. Lenz (2003), 43.

17) Cf. Hulser (1992), 17-34.

18) See Augustine (1973), esp. c. 2.

19) See Cameron (2004); cf. Rosier-Catach (2004).

176

M. Lenz / Vivarium 45 (2007) 169-188

Yet, the transmission of the relevant doctrines to Abelard and to the Albri-
cani cannot be taken as matter of simple adoption. We can see this in the view
taken by the generation after Abelard. Already the Dialectica Monacensis —a
text dating from around 1160 (i.e. not more than thirty years after Abelard
and Alberic taught in Paris)—reveals a striking awareness of the fact that the
early 12th century version of the distinction between complete and incom¬
plete sentences, although already present in Boethius, was considered an inno¬
vation in its own right in the early medieval artes , developed independently of
the works of Aristotle and Priscian by authors such as Abelard: Et notandum
quod ista divisio orationis neque habetur a Prisciano neque ab Aristotile y sed solum
magistralis est. 2Q So, roughly speaking, we might distinguish two different peri¬
ods of discussion with respect to this notion of sentence: firstly, the debates in
the Glosule super Priscianum which mainly attempt to distinguish the gram¬
matical accounts of the doctrine from the logicians’ views; secondly, the views
of Abelard and his contemporaries, who—although referring to the Glosule —
try to combine logical and grammatical aspects. As will become clear, how¬
ever, Abelard and the Albricanus have quite different ways of combining these
aspects. But before we move on to the details of the controversy, let me briefly
recall the main points of Abelard’s position. 21

Abelard’s Position

Abelard united the traditional theses into an independent account and devel¬
oped a structural model of sentences and thoughts that amounts to a distinc¬
tion between propositional attitudes and propositional contents. And here,
unlike in Searle, the notion of completeness (perfectio ) does not only carry
a lot of theoretical weight, but is examined in detail. 22 The gist of Abelard’s
account can be resumed as follows: (1) A sentence signifies an inherence—i.e.
the propositional content ( intellectus )—, but the signification (. significatio ) of
a propositional intellectus is a property of complete sentences (such as a man
runs’) and of incomplete sentences (such as a running man’) alike. This is why

20) Dialectica Monacensis , in De Rijk (1967), 467.

21) A more detailed account of the following is given in Lenz (2005).

22) In defence of Searle, one could argue that he ascribes the expressive function not to sentences
themselves, but rather claims that the speaker expresses a proposition (in the modern sense) by
using a sentence; thus Searle rids himself of the need to explain in what way a sentence is com¬
plete. See Searle (1969), 29.

M. Lenz / Vivarium 45 (2007) 169-188

177

signification is no criterion of completeness. So a sentence is complete (2) only
if it contains a finite verb by means of which (3) the content is expressed in
accordance with a certain mental attitude {affectus animi) towards the content,
e.g. a wish (‘If only the man would run) or a belief (‘The man runs’). The
main point of Abelard’s argument, however, revolves around his new interpre¬
tation of what I call the suspense-criterion. According to Boethius and many
of his interpreters the essential criterion for the completeness ( perfectio) lies in
the assumption that, on hearing a complete sentence, a hearer does not expect
to hear more.

Abelard rejects the standard view of the suspense-criterion and refines it
with regard to his account. According to him, this criterion cannot be fulfilled
by adding to the content of the sentence, but only by expressing the proposi¬
tional attitude towards the content. So, it does not really matter whether the
hearer wants to hear more about the content; the sentence is complete if the
attitude is expressed. In the case of a sentence such as “A man runs” I might
well want to hear who precisely the man is or where he is running, but my
suspense cannot really be released by any more information—unless we would
then want to invite the possibility of an infinite regress here. Perfection is
gained rather through the expression of the attitude towards the content. So
Abelard finally propounds a model of sentences and thoughts that generally
allows for an analysis into two sentential and mental components: proposi¬
tional attitude and content. In sum we can distinguish between the following
three kinds of sentences:

(1) An incomplete sentence (... me legere, ego legens) signifies a thought (that
I read): the hearer remains in suspense; no truth value.

(2) A complete non-assertive sentence (Utinam legerem) signifies a thought
(that I read) and expresses it according to a mental attitude (an affectus
animi in optative mode): the hearer rests; no truth value.

(3) A complete assertive sentence (Ego lego) signifies a thought (that I read)
and expresses it according to a mental attitude (an affectus animi in indica¬
tive mode): the hearer rests; truth value (with respect to the dictum).

The Critique of the Albricanus in the Glossae \doctrinae sermonum 9

Abelard’s position might seem fairly coherent, but it implies quite a number
of systematic difficulties which can’t be considered here. Still, my talk of a
distinction between propositional contents and attitudes may seem rather

178

M. Lenz / Vivarium 45 (2007) 169-188

anachronistic with regard to the medieval notion of mind ( animus ) and
soul ( anima ). Since, what I call “propositional attitudes”—i.e. the affectus ,
which are partially attributed to the souls of irrational animals as well as to
humans—was quite often excluded from the analysis of sentences (and the
related mental acts). By contrast, in modern philosophy of mind the advan¬
tage of the talk of propositional attitudes is precisely that they can be generally
ascribed to the cognitive realm, and thereby allow for functionalist explana¬
tions of mental processes. Accordingly, I shall have to consider the objection
whether my talk of propositional attitudes in Abelard does not muddle up two
realms that were separated in medieval thought. In that case I would have to
withdraw my interpretation as anachronistic.

However, the objection against mixing up two distinct realms in the notion
of the propositional attitude has already been raised by the anonymous author
of the Glossae doctrinae sermonum. The Albricanus accuses Abelard of mixing
together affective and cognitive elements in the notion of affectus (or passio).
As has been pointed out already, he explicitly refers to Abelard and he sets out
his critique with the question as to how Abelard distinguishes between asser¬
tive and non-assertive sentences—but his reference to Abelards text opens
with a blunder:

However, Master Peter claims that the same truth or falsehood and the same thought is
signified by deprecative and imperative sentences and such like as well as by assertive sen¬
tences, so that the same thought is signified by the sentence “I want to read” and by the
sentence “If only I read”. Nevertheless he concedes that the former one, viz. ‘I want to read’,
is an assertive sentence, because it signifies and states the truth, whereas the latter one, viz.
‘If only I read’, is not, because—although it signifies a truth or falsehood—it does not state
it. Since [a sentence] cannot state unless by affirming or negating.’ 23

As Klaus Jacobi, Christian Strub and Peter King have pointed out, the anony¬
mous author obviously “misrepresents” Abelard’s view, because Abelard never

23) The Perihermenias-Commem&ry Glossae \doctrinae sermonum is preserved in MS Paris, B.N.
Lat. 15.015 (ff. 180ra-199ra) and partially edited in De Rijk (1966), 1-57. See p. 47: ‘Dicit
tamen Magister Petrus idem verum vel falsum et eundem significari intellectum a deprecativis et
inperativis orationibus et consimilibus et ab enuntiationibus, ut idem intellectus significatur ab
ista: ‘ volo legere et ‘ utinam legerem . Alteram tamen tamen concedit propositionem esse, scilicet
l volo legere\ quia verum significat et enuntiat, alteram non, scilicet ‘ utinam legerem , quia, licet
verum vel falsum significet, non tamen enuntiat. Enuntiare enim non potest nisi affirmando vel
negando.’

M. Lenz / Vivarium 45 (2007) 169-188

179

claimed that non-assertive sentences signify truths or falsehoods. 24 And the
Albricanus does indeed overlook a central tenet of Abelard: although Abelard
claimed that assertive and non-assertive sentences can signify the same thought ,
he did not claim that this thought is then a true or false thought. Since, on
his account, the truth or falsity of a thought is not decided with regard to
signification, but with regard to the dictum , i.e. with regard to how things are
or aren’t in reality.

I would like to argue, however, that the Albricanuss misrepresentation is an
understandable one and even a productive one. This assumption can be sup¬
ported if we recall the Boethian tenet that is discussed in this dispute. The
Albricanus clearly has Boethius’s definition of the assertive sentence in mind:
‘propositio est oratio verum falsumve significans .’ 25 It is the very definition that
inspired Abelard’s reasoning in the passage quoted by the Albricanus. If we try
to bring Boethius’s definition into line with Abelard’s tenets, then we arrive at
the following conclusion:

PI: Assertive sentences signify something true or something false. (Boethius)
P2: Non-assertive sentences signify the same as assertive sentences. (Abelard)
C : Therefore, non-assertive sentences signify something true or some¬
thing false.

The conclusion (C) is the very starting-point of the anonymous author’s cri¬
tique in the passage quoted above. Now, Abelard would have had to arrive at
the very same conclusion—which is why the “misrepresentation” is indeed an
understandable one. He would have had to arrive at this conclusion, had he
not introduced the distinction between “signifying” {significare) and “stating”
(i enuntiare ). This distinction allows for the determination of truth and false¬
hood with regard to how things are, to the dictum. So, for Abelard, Boethius’s

24) Jacobi et al. (1996), 28. I do not, however, agree with the claim they make in asking
rhetorically: ‘After all, how can an expression signify a truth or falsehood without stating it as
well?’ It is perfectly plausible that a sentence signifies a truth without stating it—namely in a
potential manner, as, e.g., in the case of certain conditionals. And the idea of potential signification
was widely discussed already in the Glosule super Priscianum and by students of William of
Champeaux; cf. Anonymus, In Periherm., MS Paris B.N. reg. lat. 13368, f. 225rb: ‘Sed viden-
dum est, quod significativum duobis modis accipi potest, scilicet vel actualiter vel potentialiter,
id est pro potens generare intellectum vel generans intellectum actualiter.’ (226va) The passage is
quoted in Cameron (2004), 101.

25) Boethius (1860), 1174B. Cf. Marenbon (2004a).

180

M. Lenz / Vivarium 45 (2007) 169-188

definition is acceptable only in the light of this distinction, at the heart of
which lies the distinction between propositional attitude and content. The
central criterion is, then, not the signification but the propositional attitude
of stating, mirrored by the, indicative mode of the verb. And this is a point
which we find adopted in later treatises again. In the Dialectica Monacensis ,
for instance, Boethius’s definition is refined accordingly: “ propositio est oratio
verum velfalsum significans indicando.” 26

The Albricanus , however, does not resort to Abelard’s solution; he declares
that the distinction between significare and enuntiare does not explain any¬
thing. 27 But if we take a closer look at the reasons that drive his argument,
we will see why the misrepresentation’ is not only understandable but even
productive.

Perspectives of Signification

Besides the author’s intention to solve the matter at hand, we must not over¬
look the wider intention of the early twelfth-century logicians, namely the
intention to establish a coherent account of the notion of signification ( signi -
ficatio ). The prevailing understanding of that term can be paraphrased with the
following definition: “to signify x (to someone” means “to cause a mental act of
understanding x in someone” or “to cause a thought of x in someone”. The

26) Dialectica Monacensis, in De Rijk (1967), 468 (italics mine). The combination of the different
definitions of the sentence and of the assertive sentence had led to confusion regarding the coher¬
ent classification of linguistic units even before Abelard’s time. Abelard’s idea that incomplete
and complete sentences can share the same propositional content, was an innovation that still
had to be defended against critiques (such as the one by the author of the Glossae \doctrinae
sermonum ’), especially when it came to the question of how to integrate Boethius’s definition
according to which propositio est oratio verum vel falsum significans. In the Dialectica Monacensis
(ibid.), however, we find that, by the time of 1160, the idea had obviously received an approval
absent from earlier texts: ‘Sic ergo describitur propositio: propositio est oratio verum vel falsum
significans indicando. ‘ Oratio pontur pro genere. Per hanc vero differentiam ‘ verum vel falsum
significans excluduntur alie orationes, que on significant verum vel falsum. Hoc autem dico
4 indicando apponitur propter huiusmodi appellationes ‘ Sortem currere, ‘Platonem disputare, que
licet verum vel falsum significent, non tamen cum indicativo modo. Hie debes notare quod hanc

differentiam ‘ indicando nec ponit Aristotiles neque Boethius_Obmittunt autem illam par-

ticulam ‘ indicando ’, quia appellationes quas nos excludimus mediante ilia differentia ‘ indicando
a propositione, quantum ad eos excluduntur per hoc nomen 'oratio, cum uterque eorum hoc
nomen ‘ oratio restringat ad intentionem perfecte orationis.’

27) See Glossae 'doctrinaesermonum, in: De Rijk (1966), 47.

M. Lenz / Vivarium 45 (2007) 169-188

181

semantics of the twelfth century often focuses on the hearer’s perspective, which
is why signification is explained in terms of the hearers understanding. 28

Particularly with respect to a coherent notion of signification, it is quite
appropriate to construe a sign as something that is given to others , something
that is rather given for the sake of others than for ones one sake. 29 Accordingly,
one could take the most disparate things as signs, such as units of language
(taken as conventional signs), the cries of the sick ones or the bark of an irra¬
tional dog (taken as natural signs). This way one did not have to ascribe a
thought to a dog’s soul, if one wanted to argue that the dog barks without
thinking and yet signifies to us (rational beings) that it is angry. Without taking
the hearer’s perspective, however, we cannot ascribe meaning to a dog’s bark.

In addition to the hearer’s perspective—which is, according to the Glosule ,
stressed very much by Boethius and which is indeed of major importance with
regard to his suspense-criterion—, there are two other complementary per¬
spectives relevant to signification. 30 On the one hand there is the notion of
signification in accordance with the original imposition or institution (often
referred to as the communis causa of the significatio), which is particularly
relevant to the distinction of grammatical units: incomplete sentences, it is
argued for instance, can be called incomplete because their grammatical struc¬
tures do not constitute independent units of language insofar as they have not
been instituted to be uttered in isolation. 31 This also makes clear that these
logicians were fully aware of the point that phrases such as homo legens (“man
reading”) are not units in the actual use of language but rather the result of
semantic analyses.—On the other hand there is of course the speaker’s per¬
spective or the so called “expression semantics”, which is associated with
Priscian and the Glosule.

28) See Marenbon (1997a), 182, who paraphrases this definition given in Spade (1980b).

29) I am very grateful to Katherine Tachau for discussing this point with me.

30) See on this Meier-Oeser (1997), 34-50. It is, however, somewhat unfair to attribute to
Boethius a ‘hearer-focused semantics’, as the author of the Glosule does. Boethius’s account of
what he calls the orandi ordo does construe communication from the perspective of the hearer
and the speaker. Cf. Cameron (2004).

3,) See e.g. Peter Abelard (1927), 373: ‘Quantum ergo ad causam inuentionis imperfecta est
oratio ‘Socrates currens’ quia, licet et intellectum constituat per se etiam prolata, non propter
hoc, ut per se hoc faceret, instituta fuit sed cum aliis uocibus iuncta. Unde et ille qui audit earn
proferri et scit earn non esse inuentam propter hoc, ut simpliciter proferretur, aliam exspectat
uocem...’ I am very grateful to Klaus Jacobi for kindly facilitating a copy of the edition of
Abelard’s Logica commentary on De interpretatione that he and Christian Strub prepare for pub¬
lication in the Corpus Christianorum. References are to Geyer’s edition.

182

M. Lenz / Vivarium 45 (2007) 169-188

Hearer and Speaker

Now, Abelards semantics is commonly associated with Boethius’s hearer-
focused account, but his new interpretation of the suspense-criterion does in
fact shift the perspective to the speaker-focused account. Since, although the
completeness of a sentence is traditionally characterised by the hearer s resting,
Abelard construes completeness with regard to the expression of the proposi¬
tional attitude. And these tacit shifts from one perspective to the other are
made explicit by the Albricanuss criticisms. Lets recall the conclusion that the
anonymous author had to draw:

P1: Assertive sentences signify something true or something false. (Boethius)
P2: Non-assertive sentences signify the same as assertive sentences. (Abelard)
C : Therefore, non-assertive sentences signify something true or something
false.

Since the author wants to avoid the conclusion just as much as Abelard, but
since he does not accept Abelard s distinction between significare and enunti -
are , he has to explain the role of non-assertive sentences in a different way: he
does of course accept PI. So he has to refute P2, if he wants to avoid C. And
this is exactly what he attempts to do, when he explains his position from the
speakers perspective:

We claim, however, that imperative or deprecative sentences and such like do by no means
signify something true or something false, but that they only constitute certain passions of
the mind. We do not concede [that they constitute] a true or false thought. For who says “If
only I read!” does not manifest a thought of the mind, but solely the will of the mind. 32

In the next part of the argument he shifts the focus to the hearer:

However, in the hearer it [viz. the sentence] generates a true thought, since the hearer grasps
through his [viz. the speakers] words that he wants to read,} 1

32) Glossae doctrinae sermonum\ in De Rijk (1966), 47: ‘Nos autem dicimus orationes imperati-
vas vel deprecativas et huiusmodi nullatenus verum vel falsum significare, sed tantum quasdam
animi passiones constituere, non verum vel falsum intellectum concedimus. Qui enim dicit
“Utinam legerem!” nullum animi intellectum sed solam animi voluntatem manifestat...’

33) Glossae 'doctrinae sermonum, in De Rijk (1966), 47 (italics mine): \.. in auditore tamen verum
generat intellectum; ex verbis enim eius concipit auditor ilium ttelle legere .’

M. Lenz / Vivarium 45 (2007) 169-188

183

The whole passage can be resumed as follows: while the speaker solely mani¬
fests his will or his affective stance without any access to the (intentional)
cognitive content of his will, the hearer does not only grasp the content of the
wish (that he reads), but also the passion (the fact that the speaker wishes it) in
a reflexive way. I call this way of grasping “reflexive”, because it is not solely
directed at a content but combined with the awareness about the way in which
we (or, in this case, the speaker) are intentionally related to the content; and
thereby the originally affective stance of the speaker is conceptualised in the
hearer’s understanding, or grasped per modum conceptus} A The following table
should illustrate the author’s position:

Sentence uttered

Speaker’s mind

Hearer’s mind

“Volo legere”

intellectus verus

—> intellectus verus

(I want to read)

“Utinam legerem!”

affectus

—> intellectus verus

(If only I read!)

These points are taken up again in the next section of the author’s argument:

This is why the sentence “If only I read!” is solely a sign [nota] of the will but not of the
thought; yet the sentence “I want to read” is a sign of a thought. Since it demonstrates and
signifies the thought of the speaker—but each of the two [sentences] constitutes the same
thought in the hearer. 35

This means that the sentence “I want to read” is taken as a reflexive under¬
standing of the sentence “If only I read!”, but the speaker manifests his thought
only in the reflexive sentence, whereas the hearer always has a reflexive under¬
standing. Therefore the sentences always produce true or false thoughts in the
hearer. Another example is given to rub the point in:

34) The phrases 'per modum conceptus and per modum affectus do not appear in these early
medieval texts, but they aptly express what is meant here. Cf. on the development of this termi¬
nology Nuchelmans (1988). A typical example where reflexive thinking is even required is the
case of lying, because here it is not enough to grasp the content of the lie, it is also necessary to
know that I am lying.

35) Glossae “doctrinae sermonum \ in De Rijk (1966), 47: ‘Unde ista oratio “Utinam legerem!”
solius voluntatis et non intellectus nota est, ista vero “Volo legere” intellectus nota est. Intellectus
enim proferentis demonstrat et significat—eundem tamen intellectum utraque in auditore
constituit.’

184

M. Lenz / Vivarium 45 (2007) 169-188

Likewise, if a boy asks that bread should be given to him, by saying “Give me bread!”, he
does not use this form of words to manifest the thought that he could have, since he does
not think that he wants the [bread], but rather to indicate this affection of the soul. Never¬
theless this form of words does constitute a true or false thought in the hearer. Since, who
hears this [form of words] understands that [the boy] demands that bread should be given
to him. Consequently, such sentences do not signify thoughts but certain affections of the
mind. This is why Priscian says: “The modes [of verbs] are inclinations of the mind which
indicate various of its affections.” Since, just as there are different affections of the mind,
there are different sentences to designate them: e.g. the imperative sentence [to indicate] an
order, the optative sentence [to indicate] a wish, and such like.... The bark of a dog signifies
for us, that the dog is angry, and laughter signifies that a human is happy, and these are true
or false thoughts. 36

A Communicative Approach

If we take a look at the mentioned conclusion again, we will have to say that
P1 (Assertive sentences signify something true or something false) holds true
for the speaker as well as for the hearer, whereas P2 (Non-assertive sentences
signify the same as assertive sentences) is true only with regard to the hearer,
so that C (Therefore, non-assertive sentences signify something true or some¬
thing false) is also true only with regard to the hearer.

Thus, the Albricanus strictly distinguishes between the perspective of the
speaker and the hearer. Whereas Abelard divides the sentence and the corre¬
sponding thought into the components “attitude” and “content”, and thereby
aims at a coherent compositional semantics of sentences, his opponent ascribes
these components distributively to the speaker and the hearer: the attitude is
ascribed exclusively to the speaker, the content (including the attitude in a
conceptualised form) is ascribed exclusively to the hearer. The Albricanuss
position entails, then, the claim that being related to a content always requires
a reflexive act of thinking.

36) Glossae ‘doctrinae sermonum\ in De Rijk (1966), 47-48: ‘Similiter cum puer petit panem sibi
dari, dicens “Da mihi panem!”, non ad intellectum quern habeat manifestandum, cum non
intelligit se hoc velle, tali utitur voce sed potius ad ilium animae affectum indicandum. Consti-
tuit tamen vox ilia verum vel falsum intellectum in animo auditoris. Intelligit enim qui audit
ipsum imperare panem sibi dari. Significant igitur huiusmodi orationes non intellectus sed quos-
dam animi affectus. Unde dicit Priscianus: ‘Modi sunt inclinationes animi, varios eius affectus
indicantes.’ Sicut enim diversi sunt animi affectus, sic ad illorum designationes diversae sunt
orationes: imperativa ad imperationem, optativa ad optationem, et sic de caeteris.... Latratus
enim canis significat nobis canem iratum et risus significat hominem gaudentem esse, qui intel¬
lectus veri vel falsi sunt.’

M. Lenz / Vivarium 45 (2007) 169-188

185

Accordingly, the hearer always performs reflexive acts in his or her under¬
standing of sentences, therefore there is only one option: a complete sentence
always produces a complete thought in the hearer. With regard to the speaker
there are two options: he either utters an assertive sentence and thereby mani¬
fests his thought, or he utters a non-assertive sentence and thereby manifests
his emotions or his will, but not a thought. In the latter case the speaker
has—just like a barking dog—no access to the content of his will or affection.
On this account, a human can gain this kind of access only by means of a
reflexive thought, which would include the affection as part of the cognitive
content, thus turning the affection into a cognitive content. In effect, the
Albricanus claims—in contrast to Abelard—that every thought that is pro¬
duced in the hearer of a complete sentence (be it assertive or non-assertive) is
a true or false thought.

Reflexivity is, then, also the decisive element for the completeness of the
thought or sentence: whereas Abelard claims—indeed not unlike Searle—that
completeness can be determined only by the expression of contents along with
non-representational affective attitudes , 37 the anonymous author explains com¬
pleteness with regard to the reflexive understanding of the hearer; and a
complete thought is, then, always a true or false thought. In other words:
he counters Abelard s semantics with a communicative (or even “pragmatic”)
approach. On Abelards account, the sentences “the reading Socrates”, “Socrates
reads” and “If only Socrates read!” have identical contents, they all signify that
Socrates reads: the first one is incomplete and signifies without any attitude (it
is—as has been explained above—not a unit in the actual use of language but
rather the result of semantic analyses); the second one is complete, it signifies
and expresses the content in an indicative mode (so the mental attitude is
what we would call a belief, as opposed to a desire); the third one is also com¬
plete, it signifies and expresses the content in the optative mode.

By contrast, the opponent claims 38 that the incomplete sentence “the read¬
ing Socrates” does not signify the same as the assertive sentence, since the asser¬
tive sentence—in contrast to the incomplete sentence—signifies a complete

37) This is extensively explained in Lenz (2005). Cf. Searle (1969), 29: ‘One cannot just express
a proposition [= ‘content ( intellectm)' in our terminology] while doing nothing else and have
thereby performed a complete speech act. One grammatical correlate of this point is that clauses
beginning with “that..which are a characteristic form for explicitly isolating propositions, are
not complete sentences. When a proposition is expressed it is always expressed in the perfor¬
mance of an illocutionary act.’

38) Cf. Glossae ‘doctrinae sermonum’, in De Rijk (1966), 48.

186

M. Lenz / Vivarium 45 (2007) 169-188

thought, and this means on his account a true or false thought. This point
becomes even more clear, if we recall how he would have to construe an opta¬
tive sentence. “If only Socrates read!” is, with regard to the speaker, solely a
manifestation of his desire. Yet, the hearer has the thought that the speaker
wants it to be the case that Socrates reads. Consequently, the sentences “I want
it to be case that Socrates reads” and “If only Socrates read!” produce the same
thoughts in the hearer, and these thoughts are either true or false. If my inter¬
pretation is correct, this account blocks any way to a semantic analysis of the
first order content (that Socrates reads), and, consequently, the truth or falsity
would have to depend on the first-person-authority of the speaker, and could
not be decided with regard to the actual states of affairs (that Socrates reads or
doesn’t). But my wish (“If only Socrates read!”) does of course require that
Socrates does not actually read.

As has been pointed out already, these claims amount to the result that the
speaker of such a sentence would have no access to the content of his or her
desire. Likewise, the hearer would have no access to the “conditions of satisfac¬
tion” of the speakers wish, since the hearer’s understanding is always true
or false, and, of course, the truth or falsity of such an understanding would
depend on the first-person-authority of the speaker. Having access to the con¬
tent of my desire, or, in other words, having the thought that my desire is
about, would require that I form the assertive sentence “I wish Socrates to
read”, since thereby I could conceptualise the affective state. By analogy, we
could conclude that the speaker, in making the statement “Socrates reads”,
generates the thought “He [the speaker] claims/asserts that Socrates reads” in
the hearer.

If this reconstruction is correct, then, according to the Albricanus , the
speakers “direct” expressions always generate “reflexive” thoughts/statements
in the hearer. To return to the introductory example: no matter whether your
friend tells you “If only I had enough money!” or “I fear that I haven’t got
enough money on me”, you will think “He wishes to have enough money”,
but neither of you will grasp the content “that I have enough money” as
such. And as the author makes clear in the case of natural signs such as a dog’s
bark or a human smile, your friend might as well express his emotions or
volitions through natural signs of anxiety or whatever, you will form a true or
false thought—a reflexive thought, that does not only involve the supposed
content of his desire but also the fact that he has the desire in relation to a

content.

M. Lenz / Vivarium 45 (2007) 169-188

187

Conclusions

It is time to draw conclusions. Abelards account, which combines affective
and cognitive components on the level of the sentence and thought with the
aim to establish a coherent compositional semantics, has been refuted in what
seems to be a peculiar way. But the position that emerged with the reconstruc¬
tion of the Albricanuss critique should not be dismissed easily. Apart from the
fact that the criticisms reveal some of the weaknesses in Abelards position
(such as Abelards tacit shifts from the speaker’s to the hearer’s perspective),
I would like to suggest a tentative association which might be worth further
consideration, although I hasten to add that the relevant terminology is absent
from the Glossae “doctrinae sermonum”. The overt distinction between the
speaker and the hearer is not simply a mixture of the different approaches as
associated with Priscian and Boethius in the early twelfth-century Glosule. It
also resembles what in later treatises on syncategoremata came to be known as
an utterance per modum affectus and per modum conceptus. What I call ‘direct’
and ‘reflexive’ sentences and thoughts is mirrored in this later terminology as
well as in the distinction between actus exercitus and actus significatus , which is
common still in the fourteenth century and beyond. 39

But perhaps it is even more significant that this suggestion might hint at the
differences as to how Abelard and the Albricanus used the traditional ingredi¬
ents of their accounts, and what could have been possible consequences of
these. In contrast to Abelard, the Albricanus uses the reference to Priscian’s
semantics of verbs (according to which its modes represent inclinations of the
soul) not to account for the completeness of the sentence, but to defend the
view that the speaker of a non-assertive sentence solely expresses a mental
attitude {affectus). The latter position reduces the semantics of whole non-
assertive sentences (with regard to the speaker) to the affective function that is
expressed by verbs.

But if we indeed assume that the speaker of a non-assertive sentence solely
expresses the affectus without any insight into its content, this clearly threatens
the principle of semantic compositionality. 40 Since, for a start, the compo¬
nents of a sentence could certainly not mean to the speaker what they mean to
the hearer. Now, Abelard would certainly have argued against this position,

39) Cf. note 34.

40) Jerry Fodor (see Fodor [2003], esp. 152ff.) has raised some remarkable questions about the
principle of compositionality which are well applicable to the problems discussed in the texts of
the Albricani. Cf. also Fodor and Lepore (2002).

188

M. Lenz / Vivarium 45 (2007) 169-188

since he explicitly endorsed the principle of compositionality with regard to
thoughts construed as acts of understanding a sentence. 41 Unfortunately, our
anonymous author is less clear on this point, but it shouldn’t come as a sur¬
prise that another contemporary anonymous author, who also belonged to the
school of the Albricani , attacked Abelard’s view and argued in favour Alberic
of Paris’s position, according to which thoughts or understandings of sen¬
tences are simple and not made up of parts—a position which overtly blocks
the principle of compositionality. 42

But although th zAlbricanuss position might appear rather weak, one should
bear in mind that we have a much smaller textual basis for the reconstruction
of the position of the Albricani than for the reconstruction of Abelard’s posi¬
tion. The debate on compositionality was revived from the fourteenth century
onwards, when authors such as William of Ockham, Adam of Wodeham,
Gregory of Rimini and Pierre d’Ailly discussed the structure of mental lan¬
guage. 43 Some of their arguments closely resemble those that were raised in the
early medieval discussions. But whether or not we have evidence to show direct
links between these and the late medieval debates, remains to be investigated.

41) See e.g. Abelard (1927), 339: ‘Sunt itaque tres actiones in intellectu propositionis, intellectus
scilicet partium, coniunctio vel disiunctio intellectarum rerum. Nec est incongruum, si ea actio,
quae intellectus non est, sit pars intellectus totius propositionis.’

42) See the anonymous Perihermenias-commmzntary GLossae ‘Cum plural, MS Berlin, Lat.
Fol. 624, 87va-96vb, partially edited in De Rijk (1966), 39ff. The crucial passage is (89vb-90ra):
‘Nomina et verba sunt consimilia intellectui. Ex ea auctoritate probant quidam quod intellectus
quidam sunt compositi, quidam simplices. Sicut voces quedam incomplexe, quedam complexe,
sic intellectus illi qui significantur a dictionibus sunt simplices; qui significantur ab orationibus,
sunt compositi. Magister Albericus probat nullum intellectum esse compositum. Intellectus
huius orationis: ‘Socrates est homo’ non est compositus ex hoc intellectu qui significatur ab hoc
nomine ‘homo’, quia cum ‘homo’ profertur, fit intellectus in anima et recedit, similiter intellec¬
tus significatus ab hoc verbo ‘est’ et hoc nomine ‘Socrates’. Et quia intellectus isti non sunt, ideo
nichil ex illis fit.’

43) Cf. Lenz (2003), 153sqq. and Lenz (2004).

BRILL

Vivarium 45 (2007) 189-202

VIVA

R1UM

www.brill.nl/viv

Res Meaning a Thing Thought:
The Influence of the Ars donati

Anne Grondeux

CNRS-HTL, Paris

Abstract

During the fourth century C.E. Donatus borrowed from the Greek tradition the idea of
replacing res corporalis with res in the definition of the noun, because of the usual equiva¬
lence between pragma and res. This change had important consequences, such as a new
distinction between corpus and res , as well as a new meaning proposed for the word res. This
new meaning happened to be questioned by later commentators, because masters of
grammar seemed to reject the Donatian distinction between corpus and res , and consider
that corpus is rather included inside res , before being later accepted by Sedulius and
Remigius, perhaps under the influence of the common expression res verbi. However, the
word remained poly-semic in spite of Donatus’ attempt, as can be seen in the Glosulae in
Priscianum (s. XII).

Keywords

body, thing, incorporeal, signify, thought, noun, Donatus

The following is a study about the special meaning assumed by the word res ,
understood exclusively as a “thing thought”. This meaning, which goes beyond
its common signification of a material thing, is witnessed in grammars as early
as the middle of the fourth century. To come to a clear understanding of both
the origins and later developments of this statement, it would be necessary to
carry out a careful examination of the parallels provided by grammars and
logical texts. Nevertheless in a study with aims as limited as those adopted here
one can only propose to shed some light upon a few relevant points of a theory
that involves both grammatical and logical concepts, in order to clarify the
relationship between the terminologies of these disciplines. 1

!) I want to thank for their reading and suggestions Irene Rosier-Catach, Brian Merrilees, as well
© Koninklijke Brill NV, Leiden, 2007 DOI: 10.1163/156853407X217713

190

A. Grondeux / Vivarium 45 (2007) 189-202

Res instead of res incorporalis

The Noun Signifying a Body or a “Thing” (Don. mai. II2)

As a starting point of medieval developments about this meaning of res , we
find the Ars maior of Aelius Donatus (written ca 350), which provides a
definition of the noun as a part of speech having a declension and meaning
either a body or a “thing” ( corpus aut rent) in a proper or in a common way: 2

Nomen est pars orationis cum casu corpus aut rem proprie communiterve significans. Pro-
prie ut Roma Tiberis; communiter ut urbs flumen. 3

The examples which are mentioned by Donatus only partly illustrate his pur¬
pose: “Rome” and “Tiber” are proper names, “city” and “river” are common
names, but only material things are here intended by these examples; this kind
of limitation will be noted and criticized by later grammarians. This definition
of the noun is supplemented further on when Donatus enumerates the
different types of appellative nouns, providing other examples (“man”, “earth”,
“sea” for the <appellativa> corporalia , “piety”, “justice”, “dignity” for the incor-
poralia)\ all these examples also occur in other grammatical treatises. 4 However
the way in which Donatus defines the meaning of the noun is quite unusual,
since every other Roman grammar defines this part of speech as meaning a
thing either corporeal or incorporeal ( res corporalis versus res incorporalis ): 5 the
artes which are derivative of Sacerdos’ (third century C.E.) all use the expres¬
sions res corporalis versus res incorporalis , instead of corpus versus res used by
Donatus, and only one tradition refers to seeing and touching to characterize
bodies; this is what we read for the first time in the Ars charisii (fourth century

as Leonora Lorenzetti, who kindly let me use her transcription of the Glosulae from ms. Metz,
BM 1224.

2) Note that the considerations about the meaning of nomen can be found in three places of the
Roman artes: the first one is the definition of what a nomen is (1), which takes place at the begin¬
ning of the section dealing with noun; the second one comes a bit further, where grammarians
enumerate the different types of nouns, starting with the appellatives (2); the third one, which
does not occur in every grammar, introduces the difference between three words, nomen, appel-
latio and vocabulum (3).

3) Aelius Donatus, Ars maior II 2 (1981, 614: 2-3).

4) Ibid. II 2 (1981, 613: 1-2): Appellativorum nominum species multae sunt— Alia enim sunt
corporalia, ut homo terra mare, alia incorporalia, ut pietas iustitia dignitas.’

5) About this terminology and its juridical connotations, see Buffa (1982, 7-28) and Grondeux
(2003, 35-76).

A. Grondeux / Vivarium 45 (2007) 189-202

191

C.E.), a Roman grammar independent from the Ars donati , which assumes a
link between being corporeal and being sensible:

Appellative names... are divided into two species, of which one signifies corporeal things,
which can be seen and touched, like a human, the earth or the sea; the other signifies incor¬
poreal things, like piety, justice and worthiness, which are perceived by the intellect alone
and cannot be seen or touched. 6

“Incorporeal things” are here for the very first time defined as “what cannot be
seen or touched, but only perceived by the intellect”, a most important notion
as concerns the definition of res for later grammarians. However, for our con¬
sideration the question is: from where did Donatus take the idea of replacing
res corporalis with corpus , and res incorporalis with res? We have to remember
here that res had never received this explicit meaning before in the Latin tradi¬
tion: res first means a good, a possession; its second sense is a factum y actio y
status , condicio y negotium (a sense witnessed in expressions such as rem gerere y
agere y mandare y narrare)\ its third meaning consists in fact in meaning noth¬
ing, so that haec res appears to be the exact equivalent of hoc , an empty word
which can be found in expressions such as res argentaria y cibaria y diuina y
publica y Venerea. However, Lucretius already provides an other meaning, in
the title of his book, De rerum natura\ the genitive could be translated in t«v
ovtcov, suggesting that res has here the same signification as the participle ens y
which has not yet been witnessed in Latin. 7 However we find no mention of
res meaning only something incorporeal.

The Influence of the Techne?

The Greek grammatical tradition provides a significant parallel with the
vocabulary that is to be found in the Techne of Dionysius Thrax 8 or in Apol¬
lonius Dyscolus. 9 The words used at the same places of these artes by these
grammarians are soma and pragma. But pragma is a word whose use in this
context seems very puzzling. 10 In the Stoic tradition it refers to an action or a

6) Flavius Sosipater Charisius, Artisgrammaticae libri V (1997, 193: 24 sq.).

7) Cf. Flury (1982, 33-43). For a later statement on the same question, see Augustine, Locutio-
nes in heptateuchum (1958, 3: 32): ‘Non hoc est “existens”, quod graecus dixit cbv, sed si dici
posset “essens’V

8) Cf. Dionysius Thrax, Techne 2 (1989, 50-51), and Buffa (1982). For the Techne , see Law and
Sluiter (1998) and Law (2003).

9) The passage is known by a scholion on the Techne , cf. Lallot (1989, 128).

,0) Cf. G. Nuchelmans (1973, 45 sq.); Luhtala (2000b, 110 sq.); Hadot (1980, 309-319).

192

A. Grondeux / Vivarium 45 (2007) 189-202

passion referring to a body acting or suffering, so that it is used as an equiva¬
lent for lekton. In the Stoic ontology lekta are one of the components of the
me , what may be thought but has no self existence. 11 The Stoic tradition deeply
influenced Dionysius Thrax, who defines the noun ( onoma ) as a part of speech
with declension signifying either a soma (e.g. lithos) or a pragma (for instance
paideia), whereas its Scholia all define the soma as accessible to senses, that is,
whatever can be seen or heard or touched, and the pragma as a noeton , which
can only be thought. The examples provided by these Scholia are quite
significant (“rhetoric”, “grammar”, “philosophy”, “god”, “reason”), as far as
they all testify that a complete change occurred, starting with soma versus
pragma (a body/a process) and ending with aistheton versus noeton (what is
perceived/ what is only conceived). It is true however that pragma and res have
always been understood as equivalents: for instance pragma means an event,
namely a res gesta, and in grammar the word used to translate pragma is res, as
witnessed by authors such as Quintilian, 12 Charisius 13 and Priscian. 14 But here
we may have the only occurrence where translating pragma with res was a
rather bad idea, because pragma had somehow lost its meaning of bodily action
or passion.

Defining the noun as a part of speech to mean a body or a “thing” ( corpus
aut rern) hence appears to be an innovation of the Ars Donati , linked to the
Techne written by Dionysius Thrax. 15 However “res” used for the first time
with this meaning by Donatus—namely a sense agreeing with the one implied
by the locution res gesta —has never been understood as signifying a process
expressed by a verb {pragma). Nevertheless this innovation provides a new
meaning for res , now understood by grammarians as an incorporeal thing, that
not only sets Donatus apart in the Roman tradition but also puts the Ars
Donati somewhat far from its Stoic sources: in the same way that “education”

n) Nuchelmans (1973, 47-55).

12) Quintilian Institutio oratoria 3 (1975, 6: 28) already provides the equivalence “res”, id est
pragmata, cf. ThLL X 2, 7 col. 1119, 56 sq.

13) The use of res for pragma by Charisius is also relevant, for this grammarian, who teaches in
Constantinople, provides an explicit equivalence between these two words, as well as eidoq
(thought, object of thought) as an equivalent of res in the same contexts, see Charisius (1997,
46: 18 and 460: 22). However, since Charisius is a compilator, his use of res sometimes proves
inconsistant, see Grondeux (2003).

14) Nuchelmans (1973, 48).

15) About dating the Techne , see Lallot (1989, 19-31 (s. II a.C. for chapters 1-10, s. Ill p. C.
for chapters 11-20); about Latin grammarians borrowing from the Greek tradition, see Law
(2003, 65).

A. Grondeux / Vivarium 45 (2007) 189-202

193

cannot be an instance of a Stoic incorporeal, one can see that the examples of
virtues that will be provided by Donatus’ commentators completely disagree
with the Stoic view of what an incorporeal is. 16

Prisciaris Agreement and Disagreement

The former approach to the problem by Priscian is to be found in the definition
of the noun provided by the Institutiones grammaticae II 22 (esp. 36: 29 and
59, 10), which unquestioningly borrows from Donatus. However, one passage
of his Partitiones (ad Aen. 5, 1), which were written after the Institutiones ,
provides a criticism of Donatus’ definition by confronting the definitions of
both Donatus (< corpus aut rem) and Apollonius ( corporalium rerum vel incorpo-
raliuni )} 1 One can understand this passage as meaning that res has exactly the
same sense as res incorporate, since Apollonius indeed uses pragma)* But it
seems to me that it rather has to be interpreted in an other sense, as far as
Priscian intends here to disapprove of the translation pragma-res\ my view is
supported by another passage of the Partitiones (ad Aen. 3, 1, Postquam res
Asiae...), a place in which Priscian explains that “although (quamvis) some
grammarians {quidam grammatici) used to call the incorporeals “things”,
everything indeed whether incorporeal or not can be called by the same name,
namely ‘thing’”, and he provides some examples as res Asiae (Aen. 3, 1) where
res is used for material goods, respublica, res familiaris, res uxoriaP

Despite this criticism, both main grammars used by scholars were support¬
ing the idea that a noun signifies either a body or a “thing”, given that the last
term is now supposed to mean an incorporeal. What will be their influence on
later treatises?

Various Grammatical Meanings of res (Fifth to Ninth Centuries)

Grammatical Reluctance (Commentaries on Donatus)

The purpose of the early commentators on Donatus’ Ars maior is to explain
the handbook in its own words, and this guiding principle applies especially

16) Cf. Brehier (1962, 7). For the occurrence of the same question in the difference between
three words, nomen , appellatio and vocabulum see Grondeux (2003).

17) Priscian, Partitiones ad Aen. 5, 1 (GL III, 480:33). Cf. Holtz (1981, 243).

18) Cf. Buffa(1982,24).

,9) Priscian, Ibid. (GL III, 475: 13).

194

A. Grondeux ! Vivarium 45 (2007) 189-202

to the first part of the definition ( nomen est pars orationis. ..) which is explained
by reference to a later passage, that is to say to the definition of the appellative
noun {appellativa) . Nevertheless, the terminology is borrowed from other
Roman artes , from the fifth century Servius for example, who was the first to
investigate and explain the definition of the noun:

Now the grammarians define the corporeal and the incorporeal name in this way: the cor¬
poreal is whatever is seen and touched, like a stone, the incorporeal what is neither seen nor
touched, like piety. 20

Servius uses both Ps. Probus 21 and grammars of Charisius’ group, in which he
could find a link, between being corporeal and being perceptible on the one
hand, and between being incorporeal and being imperceptible, that is invis¬
ible and untouchable, on the other hand. The same view is supported by the
fifth century Pompeius’ commentary, which relies on Servius’. 22 Cledonius’
commentary (also perhaps from the fifth century), which also relies on Ser¬
vius’, seems however quite different, insofar as he offers an original three-part
division between corporeals (which can be either seen and touched: man,
earth, sea) and two sorts of incorporeals; some of them can either be seen or
touched (sky, sun, air), whereas others can be neither seen nor touched (piety,
justice, dignity). Cledonius intends to give an accurate commentary of what
Donatus meant, an attempt which seems to be a criticism of Servius ambig¬
uous definitions:

The name signifies either a corporeal thing or an incorporeal one. Corporeal things are
those which are touched and seen, as a human being, the earth, the sea; incorporeal things
are those which are neither touched nor see, like piety, justice and worthiness. 23

Nevertheless Cledonius’ terminology is identical to Charisius’, insofar as he
speaks of res corporalis and res incorporate, and not at all of corpus and res as
Donatus did. From the seventh century on, the passage of Ars maior (II 2)

20) See also Servius, Commentarius in artem Donati (GL IV) 406: 29-30, as well as the passage
dealing with the different types of appellative nouns (429, 13 sq.): ‘Sunt nomina corporalia, quae
videmus et tangimus, ut terra lapis; sunt incorporalia, quae nec videmus nec tangimus, ut pietas
iustitia. The same topic is being dealt with by Servius, Primae expositiones Sergii de prioribus
Donati grammatici urbis Romae (GL IV, 143: 16-19). For an analysis of the sentence about the
philosophers and their different opinions, cf. Grondeux (2003).

21) Cf. Holtz (1981, 99 n.12).

22) Pompeius (Maurus), Commentum artis Donati (GL V, 137: 24-29).

23) Cledonius, Ars grammatica (GL V, 34: 26).

A. Grondeux / Vivarium 45 (2007) 189-202

195

quoted above provided the commentators with the opportunity to elaborate
on what Donatus meant when saying corpus aut rem. This new point of view
is relevant for our consideration, since commenting on Ars maior II 2 now
implies a careful study of each word of Donatus’ definition of the noun: com¬
mentators have now to explain what a “body” is and what a “thing” is. This
new approach to the text first occurs in the anonymous grammar known as
Ars Ambrosiana (first half of the seventh century):

Now a body is said to be whatever is touched and seen,... A thing however is whatever is
neither seen nor touched; thing (res) <comes> from the verb, as others <say>, ‘I think’
(reor), which is of the mind. But not every body can be a thing. 24

Although almost all commentators give a rather accurate definition of a body
(what can be seen and touched, or what can be seen or touched), they hardly
provide the same kind of definition for a “thing”. Indeed they are not satisfied
with the new grammatical meaning of “res”; this is what we can read in the
commentaries written by masters such as Julian of Toledo (680-90), 25 Tatwine
(eighth century) or Smaragdus of Saint-Mihiel (ninth century): they all prefer
to use the old expression of res incorporalis instead of res when commenting
on the word corpus of the Ars Donati in the definition of body as what can be
seen and/or touched. For instance, Tatwine provides a tripartite distinction
between corpus , corporate and incorporate : bodies are what can be touched and
seen {tangi et videri ), corporeals are what can be either touched or seen {tangi
et non videri vet videri et non tangi), and incorporeals what can be neither
touched nor seen {nee tangi nec videri ). 26

Only two exceptions can be pointed out, and the first of these is the Ars
Ambrosiana , quoted above. This treatise indicates the fact that not any “body”
can be a “thing”, which may be a misreading for the common sentence that
“not any ‘thing’ can be a ‘body’” (see below). The second exception is provided
by Erchanbert (a master of grammar whose identification with bishop Erchan-
bert of Freising has now been rejected): “ Rem intelligimus quorum substantia

24) Lofstedt, Ars Ambrosiana, Commentum anonymum in Donati partes maiores (1982, 6: 26-29).

25) Iulianus Toletanus, Ars grammatica, poetica, rhetorica 1 , 1 , 14 (1973, 12: 67): ‘Quare dixit:
corpus aut rem? Quia aut corporale erit ipsud nomen aut incorporale. Quid est corporale? Quod
videtur et tangitur, ut homo terra mare. Quid est incorporale? Quod nec videtur nec tangitur,
sed tantummodo in animo geritur, ut pietas iustitia dignitas.’

26) Tatuinus, Ars grammatica I, 23 (1968). See also a bit further at 1, 93-95 the distinction
between nouns meaning vet corpus vel corporalitatem and nouns meaning incorporalitatem.

196

A. Grondeux / Vivarium 45 (2007) 189-202

propria corporalibus oculis nequaquam cerni potest, sicut est sapientia , pietas>
aliarumque virtutum (-tern ed.) vel vitiorum nominal 17

What has been quoted here is the only definition of res as an incorporeal
which can be found in grammars based on Donatus. Masters of grammar
seem to disagree with this passage of the ars, and almost all of them reject the
Donatian distinction between corpus and res because they consider that corpus
is rather included inside res. This is a point deserving the most serious consid¬
eration, because it shows the influence of ancient grammars, such as Charisius’
and its group, and also Priscian’s Partitiones. The Ars Ambrosiana seems to be
the first commentary to challenge Donatus’ definition by drawing attention to
the fact that Donatus could have said “rem significans” instead of “corpus aut
rem significans” (a remark which seems to confirm the correction to the mis¬
reading that has been suggested above); it nevertheless tries to find a justification
for the definition provided by Donatus, alluding to Marius Victorinus 28 to
make clear that Donatus’ form might be illogical but more convenient. 29

A later anonymous treatise known as the Anonymous ad Cuimnanum also
questions the distinction provided by Donatus and answers quite clearly: a
body is always a thing, but a thing is not always a body, because “thing” (res)
includes everything, corporeal or not, 30 and this is what can be read in the
mid-VIIIth c. anonymous grammar. 31 Since Carolingian masters give no rea¬
son for their reluctance to explain and use res with the same meaning as
Donatus, we can only guess why this meaning proved to be so difficult for
them to accept. We have already mentioned Charisius and Priscian’s Partitio¬
nes, which both offer the distinction res corporalis versus res incorporalis , using
a different and traditional terminology. But Augustine and Boethius should
be mentioned too, because of their very different definitions. Augustine’s sys¬
tem as described in De dialectica V implies four terms, verbum, did bile, dictio,
res, and provides the following definition for res, which however is never
quoted in this context: “Res est quidquid vel sentitur vel intellegitur vel latet”;
and on the other hand Boethius gives the well known tripartite division res-
intellectus-vox. Hence the use of res for Greek pragma provided by Donatus
may have seemed up to date at the moment when he wrote his Ars maior.

27) Erchanbert, Erchanberti Frisingensis tractatus super Donatum (1948, 8, 2 sq.).

28) Marius Victorinus, De definitionibus (1888, 7-8).

29) Ars Ambrosiana (1982,1: 30 sq.).

30) Anonymous ad Cuimnanum (1992, 3: 43).

31) Ars preserved in the ms. Bern, Biirgerbibl. 322 (GL VIII, xlii-xliii).

A. Grondeux / Vivarium 45 (2007) 189-202

197

nevertheless it quickly became old-fashioned when new frameworks appeared,
involving a change of place for res , which was no more supposed to mean a
thing thought, but just a thing, no matter what it was, corporeal or not.
Although this chronological argument is difficult either to prove or to dis¬
prove, it is noteworthy that Papias contrasts the different meanings of the
word res , emphasizing that only grammarians use it with the sense of “incor¬
poreal”; this grammatical definition is explicitly compared with the definition
provided by Augustine in his Dialectica and with another found in Isidorus’
Origines (V 25, 2: Res sunt quae in nostro iure consistunt ). 32

Carolingian masters seem to be much more interested in two different
questions: the definition of the body (somehow in close relationship with
eucharistic problems) and also the incorporeals which can have a proper name.
A detailed analysis of the different points of view would be quite lengthy here,
but the theories touching on the last question can be summed up as follows:
the grammarians commenting on Donatus on this point all depend on
Priscian, who gives some examples of what a res with a proper name could be,
whether a goddess (such as Pudicitia) or a science set forth by a famous master
{grammatica Aristarchi for instance). These two possibilities are developed and
illustrated by Carolingian masters, who offer new interpretations of Donatus’
definition, especially by introducing the names of the angels (Michael, Gabriel
and so on) in order to replace the pagan gods. I only mention this Carolingian
discussion to make clear the radical change which will take place at the time
of the Glosulae , which are not interested at all in this sort of question.

Res as Distinct from corpus ( Commentaries on Both Donatus and Priscian)

During the ninth century, two grammarians defend Donatus’ point of view,
asserting that the true meaning of res is res incorporalis , and that when we
call res a material thing, we indeed misuse the term. It is relevant that both

32) Cf. Hamesse (1982, 101-2) for the detailed analysis of Papias’ and Giovanni Balbi’s definitions
(‘Proprie quidem res sunt quas solas estimatione et intellectu comprehendimus ut incorpo-
reae..quodlibet ens, unum de transcendentibus...; quae in nostro iure consistunt’); Osbern of
Gloucester, Derivationes (1996, R XV 16): ‘Item a reor... hec res, rei’; Firmini Verris Dictionar-
ius/Dictionnaire latin-frangais de Firmin LeVer (1994, 430b24-26): ‘Res... idest la chose —res est
quicquid vel sentitur vel intelligitur vel latet; res sunt que in nostro iure consistunt’; Dictionarius
familiaris et compendiosus. Dictionnaire latin-frangais de Guillaume Le Tailleur (2002, 335a25-
27): ‘Res... secundum Papiam .i. quicquid vel sentitur vel intelligitur vel latet; res sunt que in
nostro iure consistunt’.

198

A. Grondeux / Vivarium 45 (2007) 189-202

Sedulius 33 and Remigius, 34 the latter probably depending on the former or
on the same Irish sources, provide the same criticism of Priscians Partitio -
nes ad Aen. 3, 1 (res Asiae), asserting that res (proprie) incorporate est.

Commenting on Donatus’ Ars maior, Sedulius looks very different from
earlier commentators: his definition of corpus for instance emphasizes the ety¬
mology (the body is hence described as a jail for the heart, cor-pusP or as
something which can become rotten, corpus-corruptio) , but does no longer
insist on what a (touchable or visible) body is. 36 But Sedulius Scotus is also one
of the first commentators on Priscianus’ Institutiones grammaticae , 37 and as
such his commentary on IG II 22 needs to be discussed.

Sedulius avoids introducing in his statement the definition of res found in
his commentary on Donatus. He provides the same kind of examples that have
always been used for the res incorporates (cum * iustitiam ”, u dilectionem\ “col-
orem \ “formam” ceteraque accedentia nomino ), but this does not diminish the
obscurity surrounding this question. Later in the text he makes no further
mention of the distinction between corpus and res , and his commentary on
Priscian looks very Porphyrian (see the terminology: definitio y species , differentia).
It may be worthwhile to investigate the explanation provided by Sedulius for
each term involved in the definition of the noun, substantia and qualitas , in
that Sedulius gives essentia as an equivalent for the first one and accidens for the
second one, which is close to the solution which will be proposed by the Glo-
sulae two centuries later. 38 Sedulius here argues that the property of the name
only depends on its quality, namely that the noun has not been invented to
signify substances but qualities, and that here is to be found the difference
between noun and pronoun. 39 It is interesting to contrast his commentary with
one that has been attributed to John Scotus Eriugena, who rejects the distinc¬
tion between res and corpus , emphasizing that a body is indeed a thing. 40

33) Sedulius Scotus, In Donati artem maiorem 2 (1977, 66).

34) Remigius Autissiodorensis, Commentum Einsidlense in Donati artem minorem 16(1912, 11).

35) Cf. Grondeux and Jeudy (2001, 139-160).

36) However one should notice that a later master called Israel had an other opinion, relying on
Augustine’s, about what a res is: “Dicitur autem res secundum Israhelem quicquid sentitur vel
intelligitur vel latet (= Aug. dialect. 3). Sentiuntur corporalia; spiritalia intelliguntur; latet vero
deus et informis materia”; Jeudy (1977, 754).

37) Sedulius, In Priscianum (1977).

38) Sedulius, In Priscianum (1977, 78: 62 sq.).

39) Sedulius, In Priscianum (1977, p. 79: 86 sq.).

40) Cf. Luhtala (2000a, 115-88, ad II 22 p. 157): Omne quod est aut corpus aut res est. RES dico
incorporates creaturas. In usu namque latinae linguae omnia sensiblia et visibilia corpora dicuntur.

A. Grondeux / Vivarium 45 (2007) 189-202

199

Later Meanings of res (Eleven to Twelfth Centuries)

When the Ars maior almost disappears from the classroom, scholars still read
Priscian’s Institutiones where they can find the same kind of distinction
between corpus and res. But they no longer have access to the traditional set of
examples provided by earlier grammars ( pietas , and so on), which Sedulius
still knew.

New Sorts of res (incorporales)

The first consequence of the disappearance of Donatus’ Ars maior is that the
definition of the noun (IG II 22) proves difficult for the Glosulae to comment
on, since nobody seems to remember what grammatical incorporeals were:

A name is either proper or appellative. (Gl) Moreover (om. KI) it divides substance into
(om. KI) corporeal and incorporeal. ‘Thing is what incorporeal substances and all things in
the categories other than substance are called, ‘bodies’ are what these visible things that are
subject to the corporeal senses are called. (G2). Or ‘bodies’ is what all substantial things are
called, ‘things’ is what all accidental things are called. 41

The anonymous authors of the Glosulae are not sure whether corpora and res
should be explained in a Porphyrian or in an Aristotelian framework. The
first assumption here seems to refer to Porphyry’s tree, while the second one
seems to be connected with the Categories , the latter not very different from
the former. But grammatical incorporeals have now utterly disappeared, and
that they have been replaced by others (see below).

The second consequence is that commentaries on Priscian rely on the
terminology used by Boethius translating Porphyry’s Isagoge. Hence they no
longer need old-fashioned examples of incorporeals ( pietas ...) since words
such as communem qualitatem immediately draw their attention to univer¬
sal; 42 or when Priscian provides “homo” as an instance of a common name,

Quae autem solo intellectu noscuntur RES appellantur. Itaque ideo etiam corpora RES appellari
possunt. Omnis ergo creatura generaliter REI nomine appellari potest. Omnis erga creatura RES
dicitur, non autem omnis RES corpus.

41) Glosulae ad II22 (Koln, Dombibl. 201, s. XII = K I4ra; Metz, Bibl. Mun. 1224, s. XII = M 19vb;
Incunable edition apud Georgius Arrivabenus, Venezia, 1488 = I 26r).

42) See Glosulae ad II 22 (M 20a): ‘Sequiturque pars distribuit. .. Sensus est: nomen significat
rem subiectam vel similem aliis per aliquam communem qualitatem, velut ‘homo’ significat plures
consimiles et unitos in eiusdem convenientia qualitatis, scilicet rationalitatis et mortalitatis...’.

200

A. Grondeux / Vivarium 45 (2007) 189-202

they immediately think of another “homo”, namely the species incorporalis,
with special attention to the divine ideas mentioned in /(S’XVII 44. 43 At least
they all seem to appreciate the idea that incorporalia are always called res . 44

The third consequence is that it is again unclear whether corpus is separated
from res or not. Res still has a large scope, so that some authors cannot avoid
using it inside its own definition (cf. Gl: “res” vocat... omnes res (!) aliorum
praedicamentorum ; G2: corpora vocat omnes res substantiates, res omnes [sc. res']
accident ales\ but compare Gl ista visibilia not *res visibiles). Res indeed still
means res intelligibilis, res incorporalis , res incorporea and at the same time any
sort of thing. It even seems to have a much larger scope than before, as far as
it may sometime comprise the ten categories (since G2 provides a division of
res between res substantiales and res accidentales) . But contrasting this set of
definitions with the commentary on IG II 18 (Proprium est nominis ...)
appears as quite puzzling because praedicamenta 2 to 10 seem to play the same
part beside qualitas instead of res 45 The Glosulae say that both substantia and
qualitas are not those of the Categories: substantia has to be understood as
essentia , while qualitas has to be taken largely as if it were any accident, 46 which
constitutes a proper or common quality. This rewriting of the Priscianic
definition intends to make clear that substantia {IG II 18) (= essentia =) corpus +
res {IG II 22). But we have to remember that Glosulae ad II 22 also suggest a
division of res into res substantialis (corpus) and res accidentalis (res), which
provides a new meaning for res , namely essentia. What happens indeed is that
the species (e.g. homo), which can be called a res, since it is an incorporeal,
contains individuals, which are called (see Priscian IG II 22) corpora.

A Polysemic Word

As mentioned above, the word has remained polysemic in spite of Donatus’
effort. Res (incorporea) is for instance used to mean the proprietas (or commu-
nio) which is the quality, proper or common, namely the reason of the inven¬
tion of the name:

43) Priscian, Institutionesgrammaticae XVII 44 (GL II, 130: 11).

44) See for instance William of Conches’ Glosae in Priscianum (1981, 36 ad II 22): Iterum sub¬
stantia dividit in duo, in corpus et rem. Corpus vocat omne corporeum sive verum sive fictum,
rem vero vocat incorpoream substantiam, quia retur, id est existimatur, non corporeo sensu
percipitur.

45) Glosulae ad II 18 (K, f.l3rab; I, f. 24v-25r; M 18rb-va).

46) See also Sedulius, In Priscianum (1977, 78).

A. Grondeux / Vivarium 45 (2007) 189-202

201

If someone should ask about whiteness, since it is an appellative, what {om. Kl) common
property it designates, we say that it signifies an incorporeal thing in that it is clear and (/]
which KM) is the opposite of black; and similarly the properties are in all others {inu. /),
although they are not subject to human cognition. 47

Moreover, the polysemy of res can be observed in another passage of the Glo¬
sulae, in which the commentator intends to investigate whether homo mean¬
ing the species and not the individuals has to be considered as a common or a
proper name; in this section one can see that res successively signifies the per¬
ceptible object (Rl) which the impositor nominis had before him while giving
the name homo for the very first time, and the other homines that an animus
rationalis will be able to see and compare afterward (Rl’); then it means the
common and incorporeal property (R2) that this animus rationalis can con¬
ceive because of the similitudo existing between (Rl’); note that this animus
rationalis conceives this res (R2) as almost {quasi) existing by itself as if in the
natura rerum (R3), namely as beings (R3) are. 48

The polysemy of the word res clearly appears three centuries later in a
commentary on the Graecismus. The anonymous author here comments on
verses XII 327-330, which explain that res has to be said about invisible things,
because of its etymology ( reor ), and that visible things are called bodies, given
that any body is a thing. Commenting on these verses which only recall an old
question without providing any answer, the anonymous author also brings up
various opinions without trying to make their coherence clear. 1) Res is said in
a proper way about something that cannot be seen; 2) any body is a thing, but
any thing is not a body; 3) Res is said in a proper way de rebus intelligibilibus
because of its etymology {res a reor); however it happens to mean a corporeal
thing, as we say “give me my thing” speaking of a book; 4) different meanings
of corpus; 5) various meanings of res with justification provided by poetic,
logical and patristic quotations; 6) other meanings of res provided by three
differential verses. 49

When Donatus borrowed from the Greek tradition the idea of replacing res
corporalis with res , he may have been misled by the usual equivalence between
pragma and res. The consequence of this was a new meaning proposed for res ,
which was questioned by many commentators before being later accepted by
Sedulius and Remigius, perhaps under the influence of the common expression

47) Glosulae (K f. 15ra, M 21rb, I f. 27v).

48) See Glosulae (K I4rb-va, M 20rb).

49) Ad Graec. XII 327-330 ms. Paris, BnF lat. 14746 f. 117rb (XV s.).

202

A. Grondeux / Vivarium 45 (2007) 189-202

res verbi. This brief study should be supplemented in three directions. First, res
is at the same moment the object of an important controversy about the per¬
sons of Trinity (three res or three nescio quid), hence it would be interesting to
check out the relationship between the terminology of this controversy and
the uses of res that have been exposed here. Second, they could be compared
with the way Abelard uses the word. 50 Finally it would be interesting to con¬
trast the logical use of the word with the technical use made by later metaphy¬
sicians, whose roots are likely to be found in the grammatical and logical use,
as suggest the common etymology res a reor, res a ratus . 51

50) Cf.Jolivet (1975, 531-545).

51) Bonaventure I 25, 3 p. 446b: ‘Res enim dicitur a reor, reris, quod dicit actum a parte animae.’
See also ibid. II 37, 1, p. 876a. Alexander Hales, Glossa in quatuor libros sententiarum P. Lombardi
1 (1951, 332): ‘Res dicitur ut se ipso ens; res enim dicitur ut ens ratum.’ Quoted by Hamesse
(1982,91-104).

BRILL

Vivarium 45 (2007) 203-218

VIVA

RIUM

www.brill.nl/viv

The Logic of Being: Eriugena’s
Dialectical Ontology

Christophe Erismann

University of Cambridge

Abstract

In his major work, the Periphyseon , the ninth century Latin philosopher John Scottus
Eriugena gives, with the help of what he calls “dialectic”, a rational analysis of reality.
According to him, dialectic is a science which pertains both to language and reality.
Eriugena grounds this position in a realist ontological exegesis of the Aristotelian
categories, which are conceived as categories of being. His interpretation tends to
transform logical patterns, such as Porphyry’s Tree or the doctrine of the categories,
into a structure which is both ontological and logical, and to use them as tools for the
analysis of the sensible world. The combination of dialectic interpreted as a science of
being, capable of expressing truths about the sensible world as well as about discourse,
with an ontological interpretation of logical concepts allows Eriugena to develop his
metaphysical theory, a strong realism. Eriugena not only supports a theological realism
(of divine ideas), but also, and principally, an ontological realism, the assertion of the
immanent existence of forms. Eriugena claims that genera and species really subsist in
the individuals: they are completely and simultaneously present in each of the entities
which belong to them.

Keywords

Eriugena, categories, universal, realism

The ninth century Irish philosopher John Scottus Eriugena 1 provides one
of the rare medieval examples of a Neoplatonic approach to logic. 2 His

1) Eriugena was the translator into Latin of several works by late Greek Neoplatonic authors
(Maximus the Confessor and Pseudo-Dionysius). As a master of Liberal Arts in the Palatine
school, he was also engaged in the study of grammar and logic.

2) On the existence, during the Middle Ages, of a Neoplatonic tradition in logic, see De Libera
( 1981 ).

DOI: 10.1163/156853407X217722

204

C. Erismann / Vivarium 45 (2007) 203-218

philosophical project is inherited from Greek Neoplatonic scholasticism in
several respects. First, John assumes the legacy of Porphyry—the incorporation
of the Aristotelian doctrine of the categories 3 into Neoplatonic metaphysics.
Secondly, his method is essentially syncretic: Eriugena fits wholly into the
Neoplatonic exegetical project of harmonizing Plato and Aristotle s doctrines. 4
Before Eriugena, Boethius—in his commentary to the De interpretatione 5 —
had already stated his intention to Aristotelis Platonisque sententias in unam
revocare concordiam. Thirdly, Eriugena sees dialectic as an inquiry which is
ontological, 6 7 far more than linguistic, in nature, and which deals with the
nature of things and their principles, with the order and structure of reality.
In this, he agrees with the Neoplatonic criticisms addressed to Aristotle by
Plotinus and Proclus among others, who accuse Aristotle of ‘reducing’ dialec¬
tic to a science restricted to language and reasoning. Fourthly, the two main
texts on which Eriugena based himself in constructing his dialectical ontology
depend on the Aristotelian doctrine of categories, but have a Neoplatonic
background which affects their content. These texts are, on the one hand, the
Categoriae decern 7 —a paraphrase of the Categories originating from the circle
of Themistius—and on the other hand, Porphyry’s Introduction (Isagoge ) to
the Categories. In line with Porphyry’s exegetical project, the Isagoge follows an
Aristotelian methodology. None the less, it does convey the conceptual ele¬
ments which allow some medieval philosophers to construct an ontology
different from that of the Categories . 8 Early medieval Aristotelianism is con¬
ceived against a Platonic background.

Eriugena’s thought is also affected by the more general intellectual circum¬
stances of the early Latin Middle Ages, 9 in particular by the relatively small

3) The attempt mentioned in Periphyseon 469AB to subsume the ten categories under two
other, even higher, genera ( motus and situs) is illustrative of Eriugenas relation to the Platonic
tradition.

4) On the Neoplatonic project of harmonisation of Plato and Aristotle, see Karamanolis (2006).

5) Boethius (1880), 79:9-80:9. See Zambon (2003).

6) Cf Samstag (1929), 8: ‘Die Dialektik ist die philosophische Grundwissenschaft oder Sein-
slehre und die ontologische Methode des Johannes, der damit seine Zugehorigkeit zum Platonis-
mus sehr entschieden zu erkennen gibt’.

7) On Eriugena and the Categoriae decern , see Marenbon (1980).

8) On the presence of Neoplatonic elements in the Isagoge and their use by Eriugena, see Eris¬
mann (2004).

9) For remarks on the way in which the availability of texts can determine the understanding of
metaphysics, see De Libera (1999b). On the peculiarity of early medieval metaphysical thought
in this respect, see Erismann (2003).

C. Erismann / Vivarium 45 (2007) 203-218

205

number of philosophical works available. 10 Philosophical thought was princi¬
pally centred on the Categories and the Isagoge. This observation applies par¬
ticularly to what may be called ontology’: the rational inquiry into the nature,
structure and order of the different entities that make up the world. During a
period in which Aristotle’s Metaphysics could not be read, nor could Plato’s
dialogues such as the Republic or the Parmenides , the absence of texts strictly
devoted to metaphysical issues led philosophers to look elsewhere in order to
find appropriate conceptual tools. They found them in the texts of the Logica
vetus. This led the authors of this period to build their metaphysical thought
by using concepts which had originally been intended for the analysis of pred¬
ication and other logical purposes.

This point about the history of texts is useful for understanding our problem.
The Neoplatonic notion of dialectic is based on the idea that dialectic is a science
whose subject is primarily natural things and their principles. But the conjunc¬
tion of the two factors—the absence of strictly metaphysical texts and the presence
merely of writings, the Categories and the Isagoge , which are indeterminate as to
their subject (things or words)—was probably responsible for the lack of strict
demarcation, in the thought of several early medieval thinkers, between logic,
philosophy of language and ontology. Such an absence makes it much easier to
think that dialectic has ontological implications and that the texts of the Logica
vetus should be interpreted as teachings about things, and not only about words.
These are, as we shall see, the two pillars of Eriugena’s ontological project.

The reading of the Logica vetus given by Eriugena tends to transform logical
patterns, such as Porphyry’s Tree or the doctrine of the categories, into a struc¬
ture that is ontological as well as logical, and to use them as tools for the
analysis of the sensible world. This use is not the result of an accidental cor¬
ruption, but rather that of a reasoned, rational and philosophical approach
based on an understanding of dialectic as ontological in nature. Such a use, in
ontological reasoning, of notions imported from the field of logic, embodies
an original approach to metaphysics.

Eriugena’s Philosophical Project: A Dialectical Ontology

In the Periphyseon , Eriugena offers a rational analysis of the structure of the
world, its origin and development. The world is understood according to a

10) The only ancient philosophical works available were the fragment of the Timaeus translated by
Calcidius, the different parts of the Logica vetus , i.e. principally the Categories (or the paraphrase

206

C. Erismann / Vivarium 45 (2007) 203-218

Neoplatonic causal scheme, 11 and is analysed according to its relation to its
creator, first remaining in him in a state of perfection, then proceeding, and
finally converting into him again. This paradigm was clearly formulated by
Proclus in axiom 35 of his Elements of Theology: ‘Every effect remains in its
cause, proceeds from it and reverts upon it’. The application of this scheme by
Eriugena leads to the following result: the moment in which the created world
remains in God corresponds to the intelligible creation of the world in the
form of primordial causes or divine ideas; it is the natura creata et creans. Then
comes the procession into the sensible world, the moment of the realisation of
the division into genera and species: this is the domain to which the categories
apply. At this moment, each entity is determined by space and time. The last
stage is that of the return to the creator, the synthesis.

I shall focus on the result of procession, the sensible world as we experience
it. The ontology of the sensible world put forward by Eriugena is the philo¬
sophical centre of his doctrine. His ontology can be called dialectical for two
reasons. First, the sensible world is the domain to which the categories apply,
that of the division into genera and species. The entire discussion of the prob¬
lem of the categories is concerned with excluding God, and the primordial
causes, from the scope of the categories and predicables; God is not a sub¬
stance, nor a genus. 12 Eriugena limits the scope of the categories to sensible
realities alone, as can be seen from his inquiry into praedicatio in divinisP This
point establishes a similarity between him and Plotinus, whose intention in
the treatises 42-44 on the genera of being {EnneadsN 1,1, 2 and 3) was to show
that the Aristotelian categories do not take into account the intelligible realities.

Categoriae decern ), Porphyry’s Isagoge , the De interpretation, and Boethius’ works ( Opuscula
Sacra, Consolatio and some logical commentaries and monographs).

n) This point highlights Eriugena’s ‘Platonism’. He is in line with the Greek Neoplatonic think¬
ers who understood the structure of reality, in the words of Gersh (1978), 125, as a ‘continuous
series of causes and effects in which each term is related dynamically to the previous one;
it “remains” in its prior (manifests an element of identity with it), it “proceeds” (manifests an
element of difference), and it “reverts” (strives to re-establish the identity)’.

12) Cf Eriugena, Periphyseon 463C [I, 33: 919]: ‘Deus autem nec genus nec species est’; and
589A [II, 87: 2054-2056]: ‘Similiter neque genus est neque forma neque species neque numerus
neque OYCIA siue generalissima siue specialissima subsistit’. Passages from the Periphyseon are
quoted from the edition of E. Jeauneau. All references are to the Patrologia Latina and, in brack¬
ets, to the edition of CCCM [volume, page: lines].

13) Periphyseon book I; e.g. 463B [I, 33: 903-905]: ‘dum ad theologiam (hoc est ad diuinae
essentiae inuestigationem) peruenitur, kategoriarum uirtus omnino extinguitur’ and 463C
[I, 33: 916-920]: ‘Clare conspicio nulla ratione kategorias de natura ineffabili proprie posse
praedicari [...]. Nulla igitur kategoria proprie deum significare potest’.

C. Erismann / Vivarium 45 (2007) 203-218

207

The second reason lies in the fact that Eriugenas ontology is based on the
categories understood in a Porphyrian way, as the principles of multiplicity, 14
and on the division of substance known as Porphyry’s Tree {Isagoge 4: 21-25
and 5: 1-5). The result of Eriugenas syncretism is that the categories apply to
the sensible world and are integrated into a wider metaphysics.

Eriugena gives an analysis of the sensible world, in which universal and
particulars, general and particular entities, are combined in a hierarchy
that runs from the unique genus generalissimum down to the multiplicity of
individuals—the individuals being understood, following Porphyry {Isagoge 6:
19-24), as the numerical multiplication of a unique and common specific
essence. Eriugenas masterpiece is based on the identification of the dialectical
process of diuisio or diairetike with the ontological process of processio or
proodosP As a result, the logical operation of division corresponds exactly to
the ontological procession, that is, to the movement of creation of the world.
The logical division of ousia into genera, species, and individuals, 16 and the
arrangement into categories (among which space and time) are the determina¬
tions of the sensible world. Dialectic has two natures: it is an internal rule of
reality, or law of nature, and a paradigm for correct reasoning. The correctness
of dialectical reasoning is guaranteed by its ontological foundation. Dialectic
as a means of human reasoning is truth-attaining because the human mind
recreates a movement of division or synthesis which is present in things. 17

Eriugenas ontology of the sensible world has two main foundations: on the
one hand, the Neoplatonic understanding of dialectic, confirmed by Augus¬
tine, and on the other, the realist interpretation of the Categoriae decern —the
text through which Eriugena knew the doctrine of the categories—and of Por¬
phyry’s Isagoge. Before pursuing these issues further, it may be useful to explain
what is meant by a Neoplatonic understanding of dialectic, and to outline
the ways through which Eriugena came to be in contact with such a doctrine.

14) See Isagoge 5: 9-12 (1966 11:1) where the genus generalissimum is said to be a supremum et
primum principium.

15) This idea of assimilating the logical division to the ontological procession is typically Neopla¬
tonic. The first occurrence is probably Iamblichus (quoted in the Scholia Platonica (adSoph. 216A):
“The diairetike imitates the procession from the One”; see Bechtle (2002), 175-218. This idea is
developed by Proclus ( Theol ’ Plat. I 9); see Lloyd (1982), 22-23.

16) See Moran (1989), 139: ‘Dialectic is not just how the mind proceeds, it is also the way the
hierarchy of reality itself is ordered’.

,7) See Beierwaltes (1986), 220-1: ‘Being therefore possesses a dialectical structure that can be
adequately translated into dialectic as a methodology’.

208

C. Erismann / Vivarium 45 (2007) 203-218

Eriugena’s approach can be understood as a return to a Platonic notion of
dialectic, conceived as the science which allows an inquiry into the ontological
structure of reality. Aristotle reduced the ambitions of Plato’s conception
of dialectic, bringing it back to a science of discourse and reasoning. But Neo¬
platonic authors such as Proclus—the main inspiration of Eriugena’s Greek
sources—tried to give back to dialectic its ontological scope. Proclus defends
an objective’ dialectic, by means of which he tries to explain how the one
passes into the multiple. The subject of dialectic is the study—in things and in
thought—of the emanation (proodos) of all reality from a unique principle.

Besides Pseudo-Dionysius, the author who was instrumental in the trans¬
mission of the Neoplatonic conception of dialectic is evidently Maximus the
Confessor, whom Eriugena read and whose Ambigua and Quaestiones ad
Thalassium he translated into Latin. 18 Through his work, Eriugena came into
contact with the Greek Neoplatonic notion of dialectic. Maximus’ Ambigua x<)
transmits Porphyry’s logical or ontological ladder, which goes from the genus
generalissimum right down to the species specialissimae through the general gen¬
era, and the fundamental idea of division as a natural progression.

Augustine’s De doctrina Christiana provides a theological framework which
facilitates Eriugena’s passage from logic to ontology; according to Augustine,
logic is truth-giving because it reflects the divine order found in things. At
least in the first part of his works, Augustine develops an understanding of the
truth of logic and of dialectic, which is based on their correspondence to the
divine order of things. Far from being human inventions, they are the knowl¬
edge of the rational structure which God has inscribed in the midst of cre¬
ation. Augustine develops the theme of the divine origin of dialectic that was
already present in Plato’s Philebus (16CD), and gives it a Christian tone. The
rules of dialectic, according to Augustine, do not lie in an exterior revelation,
like that of the Scriptures, nor are they invented by man; they appear in the
nature of things, following the order decided by God for the world. 20 For
Augustine, the liberal arts were not invented or instituted by man on the basis
of a convention, but reflect the order of things established by God according
to an eternal and immutable reason. Such a doctrine has obvious consequences

18) On Eriugena and Maximus on the Categories , see Kavanagh (2005).

19) See for example Maximus, Ambigua 1177 (1988, 92: 1389-1396).

20) Augustine, De doctrina Christiana II, XXXII, 50: ‘Ipsa tamen ueritas conexionum non insti-
tuta, sed animaduersa est ab hominibus et notata, ut earn possint uel discere uel docere; nam est
in rerum ratione perpetua et diuinitus instituta; XXXV, 53: ‘Item scientia definiendi, diuidendi
atque partiendi, quamquam etiam rebus falsis plerumque adhibeatur, ipsa tamen falsa non est
neque ab hominibus instituta, sed in rerum ratione comperta. See D’Onofrio (1986a), 51.

C. Erismann / Vivarium 45 (2007) 203-218

209

for our problem. If dialectic is not a human construct, but an inquiry into the
internal rationality of reality, its subject necessarily lies outside the framework
of speech; dialectic pertains to things themselves. 21 This is one of the textual
foundations for the status which Eriugena gives to dialectic—that of a law of
nature.

Eriugenas Understanding of Dialectic

Following Augustine, Eriugena gives a special status and origin to dialectic.
This passage from the Periphyseon summarizes Eriugenas thought particu¬
larly well:

[...] that art which divides the genera into species and resolves the species into genera,
which is called dialectic did not arise from human contrivance, but was first implanted in
the nature of things (in natura rerum) by the originator of all the arts that are properly so
called, and was discovered by the wise who make use of it in their subtle investigations of
reality. 22

Two things are immediately noticeable: the foundations of dialectic must not
be sought in discourse or in concepts but in the nature of things. Dialectic has
not been invented nor artificially developed, but has been discovered through
a close study of reality. Likewise, its scope is not limited to discourse but per¬
tains more fundamentally to the make-up of things. Thus, in a Neoplatonic
way, Eriugena holds that dialectic, whose role is to combine and distinguish
the natures of existing knowable things, and to give to each thing its definition,
allows a true theoretical inquiry of things (uera rerum contemplatio) , 23

Dialectic is called ‘ratio discretionis naturarum’ (48IB) and ‘doctrina cog-
nitionis et diiudicationis naturarum’ (827A). The word ‘natura is sometimes
used by Eriugena—as in this case—to mean the state of creatures submitted
to spatiotemporal categories. Dialectic deals with the categories and their sub¬
divisions into genera and species: ‘it is the function of that part of philosophy
which is called dialectic to break down these genera [i.e. the ten categories]
into their subdivisions from the most general to the most specific, and to

21) Note the interference between Augustinian texts as to their interpretation: the reading of the
De doctrina christiana probably influenced the understanding of the other treatise ascribed dur¬
ing the Middle Ages to Augustine, the Categoriae decern.

22) Periphyseon 749A [IV, 12: 284-288]. English translations of the Periphyseon are originally by
Sheldon-Williams, in O’Meara (1987), but have been frequently modified.

23) Periphyseon 486B [I, 62: 1879-1883].

210

C. Erismann / Vivarium 45 (2007) 203-218

collect them together again from the most specific to the most general’. 24
These categories are categories of what is, categories of being in the strong
sense of the word.

Eriugena ‘raises the notion of dialectic from the simple level of a scientia
sermocinalis to that of an ontological inquiry’ (D’Onofrio 1986b, 239). Far
from producing arbitrary patterns of thought, dialectic is, according to Eri¬
ugena, an art which can be discovered in the nature of things. It is not only a
logical operation, but also the reflection of a natural process. Eriugena expresses
this idea in a remarkably powerful phrase—the naturalis ratio which governs
dialectic is nothing but the rerum necessitas (917A). 23 Etienne Gilson (1942,
207) understood this clearly: ‘division and analysis are not simply abstract
methods of arranging and disarranging ideas, but are the law of being itself The
universe is a vast dialectic ruled by an internal method.’ Through its method
of division and analysis—the logical classification and ontological hierarchy of
existing beings as species and genera—dialectic follows the ordered and hier¬
archical structure of reality. The ontological ordering of beings into genera,
species and individuals is not an artificial conjecture of the human mind,
but a transposition of the structure of reality: the logical and ontological
classifications are similar and express an identical reality.

Eriugena often insists on the natural character of dialectical movements—
on division as a process. Far from being the result of human intellectual activ¬
ity, these movements are the work of nature and belong not to the logical, but
to the ontological order:

[...] if someone, by exercising reason in accordance with that art which is called analytike ,
unites the individuals into their species and the species into their genera and the genera into
OUSIA by gathering them together, they are said to suffer. It is not because he himself
gathers them, for they are gathered, as also they are divided, by nature; but because he
seems to gather them by an act of his reason. And similarly, when he divides them, he is said
to act, and they to suffer. 26

Thus division, whose ontological nature is clear for Eriugena, is studied by
physics, a science between theology and ethics, which is ‘the science of natures
which pertain to the senses and to the intellect’ (629B). The relation between

24) Periphyseon 463B [I, 32: 900-33: 902].

25) Probably a resurgence of Gregory of Nyssa’s De hominis opificio (201B-204A); see Gregory
(1977).

26) Periphyseon 472B [1,44: 1283-1288].

C. Erismann / Vivarium 45 (2007) 203-218

211

dialectic and the structure of the sensible world, as well as the background of
Porphyry’s Tree and of the five predicables, are obvious:

But science is the power by which the contemplative mind, whether human or angelic,
discourses on the nature of the things which proceed from the primordial causes through
generation and which are divided into genera and species by means of differences and
properties, whether it is susceptible to accidents or without them, whether joined to bodies
or altogether free from them, whether it is distributed over places and times or, outside
places and times, is unified and inseparable by reason of its simplicity; and this species of
reason is called physics. 27

According to Eriugena, the objectivity and correctness of the process of dialec¬
tic is thus guaranteed: when man uses dialectical instruments for analysing the
structure of reality, he does so basing himself on a dialectic inscribed in the
nature of things.

Once the scope of dialectic is specified, it is possible to explain its two
different aspects. In his commentary on the Celestial Hierarchy of Pseudo-
Dionysius, Eriugena establishes that dialectic is subdivided into diairetike ,
which divides the highest genera down to the individuals and analytike , which
brings the individuals together into the unity of the superior genera. 28 Dialec¬
tic as the science of the distinction of genera and species’ begins always with
division. 29 This is a postulate which both Porphyry and Boethius trace back to
Plato. 30 For a philosopher who accepts the primacy of division, this operation

27) Periphyseon 629AB [III, 17: 438-443].

28) Expositiones 184C-185A (1975, 106: 578-107: 587): ‘Due quippe partes sunt dialectice dis¬
cipline, quarum una DIAIRETIKE, altera ANALYTIKE nuncupatur. Et DIAIRETIKE quidem
diuisionis uim possidet; diuidit namque maximorum generum unitatem a summo usque deor-
sum, donee ad indiuiduas species perueniat, inque eis diuisionis terminum ponat; ANALYTIKE
uero ex aduerso sibi posite partis diuisiones ab indiuiduis sursum uersus incipiens, perque eos-
dem gradus quibus ilia descendit, ascendens conuoluit et colligit, easdem que in unitatem maxi¬
morum generum reducit [...].’

29) C $ Periphyseon 628C [III, 16: 418-420]: ‘Quicunque enim recte diuidit a generalissimis debet
incipere, et per generaliora progredi, ac sic, prout uirtus contemplationis succurit, ad specialis-
sima peruenire’. The species specialissimae, in Eriugenian terminology, are the individuals.

30) Unlike the Aristotelian, who sees only abstraction, and for whom the starting point is sensible
particulars (Anal. Post. 83b; Cat. 2a 16-17), a philosopher who prefers the Platonic forma mentis
will start with the universal which divides into particulars (Philebus l6cd). Porphyry, Isagoge
translatio Boethii (1966, 12: 9-12): ‘Quapropter usque ad specialissima a generalissimis descen-
dentem iubet Plato quiescere, descendere autem per media dividentem specificis differentiis.
Boethius, In Topica Ciceronis commentaria I, PL 64, col. 1045B2-7: ‘Plato etiam dialecticam
vocat facultatem quae id quod unum est possit in plura partiri, veluti solet genus per proprias

212

C. Erismann / Vivarium 45 (2007) 203-218

is far more than a conceptual explanation of the division of a generic entity
into its genera and species. It is a causal explanation of the being of individu¬
als: the divisive pattern carries a causal meaning.

This orientation of dialectic towards the problem of genera and species
shows the underlying direction of Eriugenas thought. For him, dialectic takes
ousia as the original constitutive principle of reality and unfolds its successive
divisions. One text highlights particularly well the privileged relation between
dialectic and ousia :

Does not that art which the Greeks call “Dialectic” and which is defined as the science of
disputing well, concern itself with OUSIA as its own proper principle, from which every
division and multiplication of those things which that art discusses takes its origin, descend¬
ing through the most general genera and the genera of intermediate generality as far as the
most special forms and species, and also never ceasing to return, according to the rules of
synthesis, by the same steps by which it descended until it reaches that same OUSIA from
which it issued, does not cease to return to it; and it yearns to rest in it forever, and to be
gathered in with it in a movement that is entirely or mostly intelligible? 31

The fact that ousia is the principal subject of dialectic shows its strong link to
ontology. For Eriugena, dialectic is first of all the science of essence. The pri¬
macy of ousia in dialectic is a postulate inherited from Middle Platonism, a
philosophical current which strongly influenced Porphyry, 32 who himself is
one of Eriugenas sources. Take, for instance, Alcinous, according to whom the
task of dialectic is firstly to examine the essence of all things and to search the
nature of each thing by division and definition. 33

Eriugenas philosophical position is based on the identification of the dia¬
lectical movement of diairetike and analytike with the ontological movement
of proodos and epistrophe. The dialectical divisio overlays the ontological processio.

differentias usque ad ultimas species separari, atque ea quae multa sunt, in unum generum rati-
one colligere.’

31) Periphyseon 868D-869A [V, 14: 360-370].

32) See Zambon (2002), 293-338.

33) Alcinous (1993), 8: 23-32: ‘Dialectic, according to Plato, has as its fundamental purpose first
the examination of the essence of every thing whatsoever, and then of its accidents. It enquires
into the nature of each thing either ‘from above’, by means of division and definition, or ‘from
below’, by means of analysis. Accidental qualities which belong to essences it examines either
from the standpoint of individuals, by induction, or from the standpoint of universal, by syl¬
logistic. So, logically, dialectic comprises the procedures of division, definition, analysis, and in
addition induction and syllogistic’.

C. Erismann / Vivarium 45 (2007) 203-218

213

Eriugena sees the diairetical descent from the universal to the particular as
analogous to the ontological relation between them. In the same way as the
mind conceives the division of a genus into its species, in nature, the genus
unfolds in order to create its subdivisions by transmitting its being. The pres¬
ence of the genus animal in this cat is the ontological cause of its animality.
The most universal entity causes the being of those which are inferior, and this
applies from the top to the bottom of the ontological hierarchy constituted by
Porphyry’s Tree.

The creation of nature is seen as the unfolding of the logical hierarchy from
the genus generalissimum down to the individuals. This Tree is presented as
being the divine plan and is called upon by Eriugena in the exegesis of certain
verses from Genesis (cf Periphyseon 748CD). Eriugena thus assigns to dialectic
the task of reconstructing in thought and language the movement of reality,
i.e. the multiplying procession of the one into the multiple and the unifying
conversion of the multiple into the one. It is then easy to postulate the exact
coincidence between notions and the substance of things. 34

A Realist Exegesis of the Categories

Eriugena gives a realist ontological exegesis of the texts of the Logica vetus : he
maintains that the subject of the Categories and of the Isagoge is primarily
things, and not words or concepts. In the words of the Neoplatonic school, for
Eriugena, the skopos of the Categories is things. Convinced that the scope of
dialectic covers the fields of philosophy of language, metaphysics, and in a
restricted manner, physics, Eriugena makes dialectic into the means of inquiry
and analysis of the ontological structure of reality, of its reflection in our
thought, and of the way we express it through language. Two claims follow
from this:

1. The categories are the highest genera of being. They are the principles of
reality.

2. Porphyry’s Tree corresponds to the ontological structure of reality. The
world is a realization of this division of substance.

34) See Periphyseon 769A [IV, 41: 1093-1095]: ‘Quid ergo mirum, si rerum notio, quam mens
humana possidet, dum in ea creata est, ipsarum rerum quarum notio est substantia intelligatur’.

35) On Eriugena on the categories, see von Perger (2005).

214

C. Erismann / Vivarium 45 (2007) 203-218

The ten categories are no longer considered as a way of classifying predicates,
but become a complete classification of things', ; no created reality can avoid cat¬
egorization:

For since, according to Aristotle, there are ten genera of things (decern genera rerum) which
are called categories, that is, predicaments, and we find that none of the Greeks or the
Latins oppose this division of things into genera [.. .]. 36

Aristotle, the shrewdest among the Greeks, as they say, in discovering the way of distin¬
guishing natural things ( naturalium rerum), included the innumerable variety of all things
which come after God and are created by Him in ten universal genera which he called the
ten categories, that is, predicaments. Nothing can be found in the multitude of created
things (in multitudine creatarum rerum) and in the various motions of minds which cannot
be included in one of these genera. 37

There are ten genera of things (decern genera rerum) which are called “categories” by the
Greeks, and “predicaments” by the Latins: substance, quantity, quality, relation, position,
state, place, time, action, passion. 38

The categories are categories of being in the strong sense. They are the highest
genera of things and not only logical entities. This conception of the categories
is associated with the claim that genera and species have real being. For some¬
one who holds that the subject of the Categories is things, this determines, at
least in part, the answer given to the problem of universal, in a realist direc¬
tion. This is precisely Eriugena’s choice.

Secondly, the structure of Porphyry’s Tree is not only a logical structure of
thought, but also the exact reproduction of reality. It is the structural pattern
according to which reality was created by God, exists in the sensible world,
and is thought by man. According to Eriugena, the world is a complete real¬
ization of Porphyry’s Tree; the hierarchy of genera and species is real . Genera
and species are not only logical categories, but ontological degrees of reality.
General ousia is the cause of the being of all its subdivisions, genera, species
and individuals (cf 605AB: Omnis substantia a generali essentia defiuit ). 39 The

36) Periphyseon 507C [I, 90: 2807-2810].

37) Periphyseon 463A [I, 32: 887-893].

38) Annotationes in Marcianum , John Scottus Eriugena (1939), 84: 3-6.

39) See the example of fire, which, being a substance, cannot have another cause than the general
essence (generalissima essentia ; 603B); and Periphyseon 750A [IV, 14: 328-332]: ‘Est enim gener-
alissima quaedam atque communis omnium natura, ab uno omnium principio creata, ex qua
ueluti amplissimo fonte per poros occultos corporalis creaturae ueluti quidam riuuli deriuantur,
et in diuersas formas singularum rerum eructant’.

C. Erismann / Vivarium 45 (2007) 203-218

215

emanation of reality from the first unique principle is ontologically parallel to
the hierarchical degrees of the dialectical division of ousia :

At the top, there is the most general genus which is called ousia by the Greeks, and essence
by us, above which it is impossible to go. For it is an essence which embraces all nature;
everything which is, subsists through participating in it, and for this reason, it is called the
most general genus. It descends through divisions, through genera and species, until [it
reaches] the most special species, which is called atomos by the Greeks, that is, the individ¬
ual: for instance, one man or one ox. 40

For Eriugena, Porphyry’s Tree is the backbone of reality. Its subsequent logical
structure reflects the nature of things. Dialectic must express in its logical
formulations the ontological priority of animal regarding man and the fact
that animal causes the animalhood of man.

A Realist Ontology

By asserting the capacity of dialectic as a science of being to express truths
about the sensible world as well as about discourse, and by giving a realist
ontological interpretation of the Categories and Isagoge , Eriugena develops his
metaphysical theory, a strong realism. His realism states the existence of uni¬
versal ante rem in the form of primordial causes or divine thoughts, and, what
concerns us more closely here, the existence of immanent universals. Eriugena
gives a clear formulation, not only of a theological realism (divine ideas), but
also, and principally, of an ontological realism, the assertion of the immanent
existence of forms.

Thus Eriugena admits into his ontology three types of substantial entities:

1) divine thoughts, which can be understood as ante rem universals;

2) immanent universals which only exist in individuals (generic or specific
essences);

3) individuals which instantiate universals.

Indicative of his syncretism, Eriugena associates a theory of divine ideas,
the medieval version of the Platonic theory transmitted by Augustine, with a
more Aristotelian realism of immanent forms. He thus accepts the existence of

40) John Scottus Eriugena (1939), 84: 10-17.

216

C. Erismann / Vivarium 45 (2007) 203-218

separated universals, which are anterior to sensible objects, and of instantiated
universals (genera and species), which subsist in individuals. Eriugena com¬
pletes his ontology with the idea of notiones , concepts obtained by abstraction,
which recreate the substance of things in the human mind and which can be
understood as post rem universals. Such a combination once again associates
Eriugena with Neoplatonic thought. Already Middle Platonists defended
immanent forms, but the combination of universals before the many, in the
many, and after the many, is a product of the Neoplatonic exegesis of Porphy¬
ry’s Isagoge by, among others, Ammonius, Elias and David. 41 Confronted with
the alternative between the primacy of the universal and that of primary sub¬
stance, Eriugena chooses that of the universal, giving priority to the model of
explanation of the singular by division of the universal; he gives only an acci¬
dental status to individuality, and ascribes substantiality only to the universal.

Eriugenas theory of divine ideas is Platonic 42 in nature, but his conception
of generic and specific universals is not, because it implies not separation but
immanence, and a complete instantiation of the universal in each individual.
Eriugena states frequently the real subsistence of genera and species in their
inferiors: they are completely and simultaneously present in each of their sub¬
divisions. For him, the species man is present one and the same, tota et simul ,
in each particular man. In the case of man, Eriugena asserts: ‘The substantial
form is [...] in all [individuals] one and the same, and in all it is equally whole,
and in none does it admit any variation or dissimilarity’. 43 This rejection of
variation in substantiality is probably inspired by the Aristotelian thesis accord¬
ing to which substance does not admit of a more and a less (Categories 3b33-
4a9; Isagoge 9: 16-23). The species is completely present in each man, but is no
less a universal one and common: ‘Ail men have one and the same OUSLA—
for all participate in one essence, and therefore, because it is common to all, it
is proper to none’. 44 For Eriugena, there are no particular essences, an essence

41) For a summary of this thesis, see Proclus (1987), 40-41: ‘Every universal, that is, every one
that includes a many, either appears in the particulars and has its existence in them and is
inseparable from them, holding its place in their ranks moving as they move and remaining
motionless when they are stationary; or exists prior to the many and produces plurality by
offering its appearances to the many instances, itself ranged indivisibly above them but enabling
these derivatives to share in its nature in a variety of ways; or is formed from particulars by
reflection and has its existence as an after-effect, a later-born addition to the many.

42) Cf Norris Clarke (1982), 116.

43) Periphyseon 703AB [III, 120: 3489-121: 3301].

44) Periphyseon 491AB [I, 68: 2088-2091].

C. Erismann / Vivarium 45 (2007) 203-218

217

is necessarily common to all the individuals of a given species. 45 The place of
the species is nowhere but in individuals, present in each of them, without
variation or deterioration: ‘human nature is distributed (difflunditur) through
all men and is wholly in all and wholly [instanciated] in each individual
whether good or evil [...], it is pure in all [...], in all it is equal’. 46 Since the
universal nature is fully instantiated by each individual, there is no difference
of degree: nullus homo alio homine humanior est (943 A).

The substance man is common to all men, who are only distinguished from
one another by a bundle of accidents particular to each individual. 47 Sub¬
stance is unique and common; individuation is accidental. 48 The individual is
substantial insofar as it instantiates its specific universal. Eriugena reduces
individuality to a bundle of universal accidents. The only thing which an indi¬
vidual possesses as it own is its body, not its essence.

The model of genus as omnipresent comes from Boethius’ second commen¬
tary to Porphyry (Boethius, 1906, 162: 23-163: 3). Boethius enumerates three
conditions which a genus is supposed to meet: it should be wholly present,
immanent in the individuals, and constitutive of the metaphysical being of the
individuals. Eriugena defends the existence of entities meeting these three cri¬
teria. He extends this model to all the degrees of Porphyry’s Tree, from the
species to ousia , the unique genus generalissimum. Thus, he describes the place
of subsistence of ousia as follows:

N.—How does it seem to you? Is not OUSIA wholly and properly contained within
the most general genera and in the more general genera as well as in the genera themselves
and in their species and again in those most special species which are called atoms, that is,
individuals?

A.—I see that there nothing else in which OUSIA can be naturally present except in the
genera and species which extend from the highest down to the lowest, that is from the most
general to the most special, that is, the individuals, and up in turn from the individuals to
the most general genera. For universal OUSIA subsists in these as if in its natural parts. 49

45) Eriugena criticizes the notion of primary substance by using the argument that, since all the
substantial being of the individual is given by the species, nothing distinguishes substantially an
individual from its species. See Periphyseon 470D-471A [I, 42:1224-1235].

46) Periphyseon 942C [V, 115: 3697-3703].

47) Here the scheme originates in Porphyry ( Isagoge 7: 19-27) and was transmitted through
Boethius’ De trinitate (2000, 167: 46-168: 63).

48) See Gracia (1984), 129-135.

49) Periphyseon 472C [1,44: 1295-45: 1305].

218

C. Erismann / Vivarium 45 (2007) 203-218

Ousia is considered the paradigmatic universal by Eriugena. A universal in the
highest sense of the word, ousia has all its characteristics—subsistence, omni¬
presence in each of its members, total immanence in its inferiors—: ‘tota enim
simul et semper in suis subdiuisionibus aeternaliter et incommutabiliter sub-
sistit omnesque subdiuisiones sui simul ac semper in se ipsa unum inseparabile
sunt’ (492C).

The omnipresence of ousia causes the subsistence of the entire hierarchy
which is subordinate to it. 50 The entity present in an immanent way in second¬
ary and primary substances can be no other than ousia , the generalis essentia at
the top of Porphyry’s Tree. 51 This realist statement also highlights the fact that
genera, species and individuals are considered equal from the point of view of
subsistence— ousia being absolutely identical in a genus and in an individual.
According to Eriugena, the universal exists in the individual as its ontological
component; its degree of subsistence therefore cannot be less than that of the
individual.

Unlike many logicians from the late eleventh and the twelfth centuries,
Eriugena is more interested in metaphysics than in the philosophy of lan¬
guage: his project is to give a rational analysis of the world. It is the structure
of reality, and not that of predication, which Eriugena wants to explain. In this
sense, he is fully representative of the early medieval philosophical project of
categorical ontology. The nature of his project itself—to give a division of
nature —is revealing. Consider, first, the idea of division : the conceptual prin¬
ciple here is that of the division of the unique principle into a multiple reality.
And the idea of nature places us immediately on the level of res. In this frame¬
work, the other theoretical elements fit coherently: the status of dialectic, the
subject of the categories and ontological realism are different aspects of a same
philosophical project. The categories are included in a metaphysics which does
not limit itself to the analysis of the sensible world, but in which the intelli¬
gible plays a leading part. In Eriugena’s work, Porphyry’s project of integrating
the categories into Neoplatonic metaphysics reaches its fulfilment. 52

50) Periphyseon 479C [I, 53: 1590-1592]: ‘Nam genera et species et ATOMA propterea semper
sunt ac permanent quia inest eis aliquod unum indiuiduum, quod solui nequit neque destrui’.
Cf Categoriae decern §57, AL I, 1-5, 145: 28: ‘ipsa autem usia genus non habet cum omnia ipsa
sustineat’.

51) See Erismann (2002).

52) I am grateful to John Marenbon for his helpful suggestions and I am indebted to his pioneer¬
ing work on the philosophy of Eriugena. I would like to thank the British Academy for support¬
ing my research.

BRILL

VIVA

RIUM

Vivarium 45 (2007) 219-237 www.brill.nl/viv

Priscian on Divine Ideas and Mental Conceptions:
The Discussions in the Glosulae in Priscianum,
the Notae Dunelmenses, William of
Champeaux and Abelard

Irene Rosier-Catach

CNRS/Ecole Pratique des Hautes Etudes, Paris

Abstract

Priscians Institutiones Grammaticae , which rely on Stoic and Neoplatonic sources, con¬
stituted an important, although quite neglected, link in the chain of transmission of
ancient philosophy in the Middle Ages. There is, in particular, a passage where Priscian
discusses the vexed claim that common names can be proper names of the universal
species and where he talks about the ideas existing in the divine mind. At the begin¬
ning of the 12th century, the anonymous Glosulae super Priscianum and the Notae
Dunelmenses , which heavily quote William of Champeaux (as master G.), interpret the
passage in the context of a growing interest in the problem of universal, raising seman¬
tic as well as ontological questions, and introducing a Platonic view on universals in
the discussions on the signification of the noun. Moreover, this same passage will be
used by Abelard to elaborate one of his opinions about the signification of universal or
common names—that they signify “mental conceptions”.

Keywords

Abelard, language, grammar, Plato, noun, signification

The interplay between grammar and logic at the turn of the twelfth century
has long been recognized. 1 A long and influential commentary on Priscian,
the Glosulae on Priscian , played an important role in the analysis of various

l) In this article, the following abbreviations are used: Abelard, LNPS = Gbssulae maestri Petri
Baelardi super Porphyrium)\ Abelard, Sup. Per. = Logica Ingredientibus: Glossae super Periermeneiasr,
Abelard, Sup. Porph. = Logica Ingredientibus: Glossae super Porphyrium , all ed. in Peter Abelard
(1919-1939); Abelard, Sup. Top. = Logica Ingredientibus: Super Topica Glossae , ed. in Peter Abelard

DOI: 10.1163/156853407X217731

220

I. Rosier-Catach / Vivarium 45 (2007) 219-237

logico-grammatical problems (like the distinction between nominatio and
signification between the existential and copulative import of the copula , the
nature and meaning of the proposition, the theory of paronyms, of the
consignificative parts of speech, of the dictum etc.), and often we find Abelard
elaborating his own views through a discussion and criticism of such analysis.
Moreover, Priscian’s Institutiones , which relied on Stoic and Neoplatonic
sources, constituted an important, although quite neglected, link in the chain
of the transmission of ancient philosophy in the Middle Ages. There is in
particular a passage, when he is discussing the vexed claim that common
names can be proper names of the universal species, where Priscian talks
about the ideas existing in the divine mind. The grammatical commentaries
on this passage, which they interpret in the context of a growing interest in
the problem of universals, both raise semantic as well as ontological ques¬
tions, and introduce a Platonic view on universals into the discussions. More¬
over, this passage will be used by Abelard to elaborate one of his opinions
about the signification of universal or common names—that they signify
“mental conceptions”.

The Notae Dunelmenses and the Glosulae

In an important article, Richard Hunt (1941-1943) called attention to the
Glosulae on Priscian and to the Notae Dunelmenses , which were then studied
by L. M. de Rijk (1962-1967), Margareta. Fredborg (1977, 1988), Onno
Kneepkens (1978, 1992) and Constant Mews (1987, 1992) in particular. The
Notae Dunelmenses (title: Note super Pris<cianum> et super reth<oricam>
veterem Tullii in fine li<bri>), are found in a Durham manuscript. They con¬
tain three sets of notes on Priscianus maior , two on Priscianum minor , and a
last one on Ciceros De inventione , attributed by Fredborg to William of
Champeaux (Fredborg 1976b, 2003). Numerous masters are quoted in the

(1954); GPma = Glosulae super Priscianum maiorem: transcriptions are made from the ms: K =
Cologne Dom B. 201, 12e th. f. lra-74rb, M = Metz, Bibl. Mun. 1224, f. Ira-11 Orb, 12e th., C =
Chartres B. Mun. 209 (248), f. l-86v, P= Paris, BnF, nouv. acq. lat. 1623, f. 1-56, and /= the incu¬
nabula version, by Georgius Arrivabenus, Venise 1488; GPmi = (1) Glosa Victorina , (incomplete)
comm, on Priscian XVII, 1-32, Arsenal 910, f. 133ra-l40vb (forthcoming edition from Marga¬
reta Fredborg), (2) complete version in three ms. B = Londres, BMBurney 238; L = Londres, BM
Harley 2713; et O = Orleans, Bibl. Mun. 90; ND = Notae Dunelmenses , D = ms. Durham, Uni¬
versity Library c. IV. 29, f. 2ra-215va; Josselin de Soissons, De generibus et speciebus, ed. in Peter
Abelard (1836); TQG = Walter of Mortagne, Tractatus Quoniam de Generali etspeciali statu rerum
universalium; Sententia de universalibus secundum magistrum R, both ed. in Dijs (1990).

/. Rosier-Catach / Vivarium 45 (2007) 219-237

221

Notae (M. Anselmus and M.A., Stephanus and M.S., M. Menegaldus, M.Ber.),
M. G. being by far the most quoted author, in sections covering Priscian
maior (ND1 and ND3) and in sections covering Priscian minor (. ND5 ).

The Glosulae on Priscian maior (GPma) themselves are preserved in several
manuscripts. Two related commentaries on Priscian minor have been studied,
which we will call in this study Glosulae on Priscian minor (GPmi ), without
any special claim about their authorship or kinship with the GPma . His author
is again a master G. (Kneepkens (1978), p. 117). The confrontation of testi¬
monies show that this master G. can be identified in many cases at least with
William of Champeaux, although the fact that “M. G” s opinions are distin¬
guished from those of the Glosulae confirms, importantly, that William is not
the author of the Glosulae (Hunt 1941-1943; Fredborg 1977; Kneepkens
1978; Rosier-Catach 2003a, 2003b, 2003c and forthcoming a and b). All
these facts show that William of Champeaux lectured on Priscian, major and
minor, probably using existing glosses on Priscian in a critical and original
way, and that Abelard knew these lectures. Often the compiler(s) of the Notae
mentions M.G. s, that is Williams, opinions as though he was in some kind of
direct contact with him, talking as if he had actually heard him teach, asking
his opinion on some point or other. 2 So he (they) may be one of the masters
whom William nominated to Notre-Dame.

Common Nouns as Proper Nouns

In three passages of the Institutiones , Priscian claims that the appellative noun
can be taken as the proper noun of a common form, or species:

[2] Priscian, Institutiones XVII, 35 (GL3, 130: 10-13). But when I say ‘what is rational
mortal animal’, I want to signify the species, that is man, which, although it seems com¬
mon to all men, is however proper to the incorporeal species. 3

[3] Priscian, Institutiones XVII, 43(GL3, 134: 19-27). They are also accustomed to join
appellatives to the infinite name ‘quis’, but more usually they join it to generic or specific
nouns, as ‘Who invented letters?’ ‘Man’; ‘What is useful to the plowman?’ ‘An ox’; ‘What
swims in the sea?’. ‘Fish’. But it seems that here the thing is an individual, when the ques¬
tion is asked in a general way about this animal itself. Similarly, when I say to many people
present, ‘Who is a grammarian among you, who is an orator, who is a doctor?’, and when

2) ND1, D, f. 8rb (ad IG, VIII, 43); ND1, D, f. 9bisrb (ad IG, VIII, 70)

3) “Cum dico vero ‘quid est animal rationale mortale?’, speciem mihi volo manifestari, id est
hominem, quae quamvis videatur esse communis omnium hominum, tamen est etiam propria
ipsius speciei incorporalis.”

222

I. Rosier-Catach / Vivarium 45 (2007) 219-257

the answer given is ‘I am a grammarian, you are an orator, he is a doctor’, <appellative
nouns> take the place of proper nouns. 4

The passage immediately following [3] is the well-known passage where
Priscian talks about the forms existing in the divine mind:

[4] Priscian, XVII, 44 (GL3, 135: 1-10). In definitions it is also customary, in order to ask
about all things which can be defined, to first posit the substantive in the neuter, and for
many appellative nouns of differentiae to be referred to it, as ‘What is an animal?’ ‘An ani¬
mated substance’, and by conversion, ‘What is animated substance?’ ‘An animal’. ‘What is
man?’: ‘A rational mortal animal’, and by conversion, ‘What is a rational mortal animal?’ ‘A
man’. The same may be done for all definitions, although, with regard to the generic and
specific forms of things, which exist in an intelligible way in the divine mind before they
come forth in bodies, they can also be proper <names> by which the genus and species of
the nature of things are shown. 5

According to the commentator, Priscian in these passages indicates that the
common noun can be used as a proper noun, for the species . The problem will
be to determine in which way the species signified can be said to be one , since
the characteristic of the proper noun is to signify a proper or singular quality.

The Universal Existing in Things and Conceived as a Singular

As might be expected, discussions on these passages will be found in the cor¬
responding glosses, in the GPmi , and in ND5. But in the GPma , the com¬
mentary on the definition of the noun (ad II, 22) mentions these two possible
uses of a common noun, as an appellative noun and as a proper noun:

[5] GPma , K I4rb-va, M 20rb-20va, C 12va, P 1 lva, [I 26v] (ad II, 22; GL2, 56: 28). And
common. Once he has given the definition of the noun, he exemplifies it through its parts,
saying that the noun is what shows a common quality of bodies, as ‘man’ shows ratio¬
nality and mortality, which are found the same in different individuals, (a) Note that when
man signifies the species itself, it is taken as a proper name, since the universal nature which it

4) “Appellativa quoque nomini infinito, quod est ‘quis’ solent subiungi, sed magis generalia aut
specialia, ut ‘quis invenit literas?’, ‘homo’, ‘quis utilis est aratro ?’ ‘bos’; ‘quis natat in mari ?’
‘piscis’; sed videtur hie quoque res individua esse, cum de ipso generaliter animali est interroga¬
te. Similiter si dicam multis praesentibus ‘quis est de vobis grammaticus, quis orator, quis med-
icus?’ et respondeatur ‘ego quidem sum grammaticus, tu orator, ille medicus’ loco propriorum
accipiuntur.”

5) “In definitionibus quoque solet ad interrogationem omnium rerum, quae definiri possunt,
neutrum substantivi praeponi et multa appellativa differentiarum nomina ad id referri, ut ‘quid
est animal ?’ ‘substantia animata’, et per conversionem ‘quid est substantia animata ?’ ‘animal’;

I. Rosier-Catach / Vivarium 45 (2007) 219-257

223

designates is understood to be like one, singular man. (b) Some people, however, state that
‘man’ is always appellative, deciding that nouns should be judged according to the nature
of their invention, not according to any mode of signification <they might happen to
have>. The person who first invented this utterance ‘man’ did not consider the unformed
species, but the thing which is subject to the senses, which he considered as sensible,
rational, mortal; and so he imposed the noun ‘man’ on the thing present to him and on
all the other things which come together in this nature. Therefore by nature ‘man is an
appellative noun. If, however, later by the intellect alone the rational soul, led by a certain
similarity with other things, conceived a certain one thing common to all men as if it
were subsisting by itself, not as it really is in the nature of things, and designated it by the
noun ‘man’, this does not seem to imply that ‘man’ should be called a proper noun, since
it designates neither a private substance, nor a private quality, but one which is common
to many. 6

Notes (a) and (b) are in fact interpolations, probably taken from glosses on
Priscian minor. They are not about the nature of universals, but about the
meaning of the appellative noun, (a) states that when an appellative noun like
“homo” signifies the species , it is taken as a proper noun, since “since the uni¬
versal nature which it designates is understood to be like one, singular man”.
Then comes the objection of quidam (b) who claim that “homo” is always
appellative: they wish to make a difference between the proper signification of
the noun, which relies on its primary invention, and a further use it may have,
a “modus signification ^”. An appellative noun was first invented to signify

‘quid est homo ?’ ‘animal rationale mortale’, et per conversionem ‘quid est animal rationale mor-
tale ?’ ‘homo’. Idem licet facere per omnes definitiones, quamvis quantum ad generales et specia¬
ls formas rerum, quae in mente divina intellegibiliter constiterunt antequam in corpora prodirent,
haec quoque propria possint esse, quibus genera et species naturae rerum demonstrantur.”

6) “Et communem. Posita nominis diffinitione exemplificat inde per partes, dicens communem
quidem corporum qualitatem demonstrans nomen, ut ‘homo’ rationalitatem et mortalita-
tem, quein diversis individuis eadem reperitur. (a) Nota quia quando 'homo ’ speciem illam
significat, ut proprium nomen accipitur, cum 6 universalis ilia natura quam designat velut unus et
singularis homo esse intelligatur. (b) Quidam tamen ‘homo’ appellativum semper affirmant, arbi-
trantes iudicandum esse de nominibus /K l4va/ secundum naturam inventionis ipsorum nomi-
num, non secundum quemlibet modum significationis. Qui autem invenit prius hanc vocem
‘homo’, non respexit ad illam speciem informem, sed ad rem sensibus subiacentem, quam con-
sideravit sensibilem, rationalem, mortalem; sicque illi presenti cum omnibus aliis in hac natura
convenientibus hoc nomen ‘homo’ imposuit. Itaque ex natura ‘homo’ appellativum est. Si autem
postea animus rationalis, quadam similitudine aliarum rerum ductus, rem quandam unam qui¬
dem et omnibus hominibus communem quasi per se subsistentem, non quantum in rerum
natura ita se habentem, solo intellectu concepit, eamque hoc nomine ‘homo’ designavit, non
tamen propter hoc ‘homo’ videtur debere dici proprium, cum nec privatam substantiam /M20va/
nec privatam qualitatem, immo communem pluribus designet.”

224

I. Rosier-Catach / Vivarium 45 (2007) 219-237

singular sensible things, with their qualities, and all the things which share the
same nature. But the mind is free to conceive this “one thing common to all
men”, the “species without forms” of man, not as it exists in natura rerum (in
the sensible things where this nature subsists), but as if it were something
subsisting by itself. 7 The mind is thus free to use the word “homo” to designate
this universal species conceived as a singular, per similitudinem , but this will
not turn the appellative noun “homo” into a proper noun. The noun keeps the
grammatical category it had when it was first imposed, and does not change
its category by a further use. Thus these quidam do not deny that the common
noun can function as a proper name, but they deny that when it does so it
should change its grammatical category. The argument is based on the distinc¬
tion between primary and secondary meaning, which is used extensively by
William of Champeaux, for instance to state that the inherence or adjacence
is signified primarily by the verb, but secondarily by the adjective (Rosier-
Catach 2003a, 2004b). So the opinion (a) that a common noun can function
as a proper name for the species as a singular can be accepted only with the
correction introduced by the quidam in (b), i.e. if one is clear that it will
always primarily be an appellative noun, and only derivatively a proper one.
An interpolation in the Chartres ms. of the Glosulae , in which the additions
often reflects Williams teaching, stresses the same idea that common nouns
are always appellative according to their first invention. 8

It is noteworthy that, except for our interpolated passage [5] above, we find
in the GPma only an immanent realist position. 9 For instance, in the well-
known gloss on the proprium of the noun, we read that there is “the same
substance or essence in all individual men”, and that the individuals are
differentiated by their “qualities” or “properties”, the signification of this com¬
mon or proper quality being thus the “cause of invention” of names. “Man”
signifies “a thing which is similar to others through a common quality”, this
quality, i.e. mortality and rationality, “being in all men which agree in it”,
whereas “Socrates” signifies this thing “as distinct from others” by a collection
of accidents “which is never as a whole identical” in other men, and thus a
proper quality. The person signified by “Socrates” is not discrete from other
men in its “substantial being”, but only by its properties, which are being
Sophroniscus’ son, being a poet, etc. 10

7) Cf. Boethius, In. Porph ., ed. II, p. 167: 14 sq.

8) GPma, K I4rb, M 20ra, C 12rb-va, P lOrb, I 26r-v.

9) For a detailed analysis and edition of these passages, see Rosier-Catach (forthcoming a).

10) GPma , K 13ra-rb, M 18va, C llva, P lOrb, I 24v (ad II, 18; GL2, 35: 6); GPma, K I4ra, M
20ra, C 12ra, P lOva, I 26r, ed. of these passages in Rosier-Catach (forthcoming a).

I. Rosier-Catach / Vivarium 45 (2007) 219-237

225

Commentaries ad XVII. 35 and XVII.43, in the GPmi , explain that an
appellative noun can be the {quasi) proper noun of the universal incorporeal
species or res , and, more specifically, that the species man is “in every man and
thus is in every one in such a way that it is every one of them”. So to the ques¬
tion “what is a rational mortal animal” the proper answer is “man”, “the com¬
mon name of all men”, and not “singular men”, 11 just as to the question “what
is useful to the plowman” the proper answer is “ox” which is “the proper name
of this universal thing which is scattered in all oxen”. 12 This idea that the uni¬
versal res is “scattered” in (or rather “through”: per) every member of that spe¬
cies echoes a passage found in a commentary on the Categories attributed to
William of Champeaux as well as in the Introductiones dialecticae secundum
Wilgelmum , 13

The GPmi , ad XVII.44 glosses further on the divine ideas:

[6] GPmi , ms. Orleans, BM 90 p. 376b (ad XVII 44 = text [4]). Although, with regard
to (133.7). Above he showed that both the nouns through which a question is put and
those which are answered are appellative, now he says that although they are appellative, yet
when they refer to generic and specific natures in their simplicity , they can be proper

n) GPmi , ms. Orleans 90, p. 373° (ad XVIL35; 130.12). “Quae quamvis. Quia dixerat quod
<ad> animal rationale mortale respondetur ilia species homo; et ita ‘homo’ istud nomen est com¬
mune vocabulum omnium hominum, unde videtur quod omnes illi responded possent, tamen
(tantum ms.) est propria id est quasi proprium nomen ipsius speciei incorporalis. Et ex hoc patet
quod species responderi debet et non singuli homines. Vel aliter supradicta potest continuari
sententia. Quia dixerat hominem speciem responderi et species est et ita est in singulis hominibus
quod est unumquodque de illis unde videretur quod omnes illi respondentur, illud removet. Con-
tinuatio litterae. Ad animal rationale mortale respondetur ilia species homo que, id est que spe¬
cies, quamvis videatur esse communis omnium hominum (130.12-13), id est quamvis sit in
omnibus hominibus et ita sit in omnibus quod sit unumquodque de illis , unde videretur quod illi
responderentur, tamen est propria ipsius speciei incorporalis id est propria species et incor¬
poralis intransitive. Et ex hoc quod species ilia est propria et incorporalis patet quod non omnes
homines respondentur sed solum propria species.”

12) GPmi , ms. Orleans 90, p. 376 (ad XVII.43; 134.19): “Appellativa quoque. Dixit superius ad
‘quis’ quandoque responderi propria nomina, quandoque pronomina, aliquando etiam appella¬
tiva nomina, nunc vero vult ostendere quod aliquando ad ‘quis’ respondentur appellativa nomina
in designatione rerum individualium. Cum vero dico quis est utilis aratro ?’ et respondeo
‘bos’, quamvis ‘bos’ sit proprium nomen illius rei universalis quod per omnes boves spargitur ,
tamen quia ‘bos’ illi interrogationi ad talem quis’ responderetur quod non convenit illi universali
rei in sua universalitate , sed tamen bos individua eius vel vox eius, et ideo respondetur in desig¬
natione individuals bovis..

13) C8 (original version), see Iwakuma (2003b), p. 327; Introductiones dialecticae <secundum
Wilgelmum> , ed. by Iwakuma (1993), pp. 58-59; cf. Tractatus quoniam degenerali, ed. by J. Dijs
(1990), p. 95, par. 6.

226

I. Rosier-Catach / Vivarium 45 (2007) 219-237

<nouns>.The letter of the text read as follows. The aforesaid nouns are appellative, and
although they are appellative they can also be proper, by which the genera and species
of the natures of things, that is of individuals, are shown, that is are signified. What he
says should be understood in this way, that just as singular nouns are proper <nouns>, so
are the aforesaid appellative <nouns>. He determines the way in which appellative <nouns>
are proper <nouns>, when he says: with regard to the generic and specific forms of
things, that is with regard to the fact that they signify genera and species. What are called
forms of things’ should be here understood as being such that from them things come into being
not formally, but materially. Which in the mind (135.8): he shows a certain property of
these forms, exist in an intelligible way in the divine mind before they come forth
in bodies, that is: God conceived them before they were brought about as existing in the
nature of thing. Or alternatively: the aforesaid nouns are said to be proper <nouns> because
they signify the generic and specific substances of things, that is the original forms which are in
the divine mind, to the exemplary likeness of which our species are made, and thus ‘man and
‘donkey’ are names for God. And the original form in the divine mind is nothing other
than God in himself, although it may seem absurd. 14

Some theses found in these passages are perfectly consonant with William' of
Champeauxs material essence theory, as reported by Abelard and others. 15
Five theses are particularly relevant for a comparison with our texts: (1) dif¬
ferent men in specie are one man; (2) there is one same universal nature or
res in all singulars belonging to the same species; (3) this universal nature is a
material essence; (4) it becomes individual through a “collection of properties”
if it is a species, through the differences if it is a genus; (5) the universal does
not exist apart from those properties and thus only exists in individual things;
(6) it can be thought in an abstracted way as being separated from any of these

14) “Quamvis quantum (135.7): superius ostendit nomina esse appellativa tarn ilia per que
queritur quam ilia que respondentur, modo vero dicit quod quamvis sint appellativa, tamen
cum ad generales et speciales naturas in sua simplicitate referantur, possunt esse propria.
Continuatio litere. Supradicta nomina sunt appellativa et quamvis sint appellativa, possunt
hec quoque propria esse, quibus genera et species nature rerum, id est individuarum,
demonstrantur id est significantur. Quod dicit, quoque sic est intelligendum, quod sicut sin¬
gulars nomina sunt propria, sic appellativa supradicta. Quomodo appellativa sunt propria
determinat quantum ad generates et speciales formas rerum, id est quantum ad hoc quod
significant genera et species. Que dicuntur forme rerum hie accipiende sunt ita quod ex eis fiant
res non formaliter, sed materialiter. Que in mente (135.8): ostendit quandam proprietatem
de illis formis que in divina mente antequam in corpora prodirent, intelligibiliter
constiterunt, id est ea Deus prius intellexit quam efficerentur in natura rei existentes. Vel ali-
ter: supradicta nomina dicuntur propria, secundum hoc quod significant generales et speciales sub -
stantias rerum, id est originales formas que sunt in mente divina, ad quarum exemplar facte sunt
nostre species , et ita ‘homo’ et ‘asinus’ sunt nomina Dei. Et nichil aliud est originalis forma in
mente divina quam Deus per se, quamvis videatur absurdum.”

15) Abelard, Sup. Porph., p. 10:17-25. Cf. C8 in Iwakuma (1999), p. 103; BIO ibid., p. 116.

/. Rosier-Catach / Vivarium 45 (2007) 219-237

22 7

properties. In GPmi (texts in footnotes 11 and 12) we find all these theses
except (3). 16 The combination of (5) and (6), associated in section (b) of text
[5] is stressed in a very close way in P3, an influential commentary on Por¬
phyry from the same period. 17 In text [6] from GPmi , there is also a clear
formulation of thesis (1) and, importantly, the second passage in italics in our
text echoes thesis (3): the “forms of things”, i.e. the genus and species, are
those from which “things are made not formally but materially”, 18 meaning
that the generic and specific forms correspond to the “matter”, and the
differentiae and accidental forms to the “form”, an idea also found in the P3
Isagoge commentary. This could be a further indication that the elaborations
introduced in the GPmi , which are not in the GPma , may be associated with
William of Champeauxs teaching.

Developments in the Notae Dunelmenses

Let us now turn to the Notae Dunelmenses. In the ND5 we find a very rich
gloss on XVII, 35, wondering about how the species can be said to be one , in
order for the appellative designating it to be called a proper noun:

[7] ND5 , f. 152rb-152vb (XVII, 35; 130.12 = text [2]). which, although it seems

COMMON TO ALL MEN, SEEMS HOWEVER TO BE PROPER TO THE INCORPOREAL SPECIES. [A] It
can be read in re and in voce. (1) An attack can be made on the statement that the question
was asked about the species as if it was one thing, and that the answer was ‘man as if it was
one thing—this, I say, can be attacked, because someone would see man being in all indi¬
viduals, and thus he could not see one thing but many, so that he could and should answer
‘Socrates’, or ‘Plato’ etc. (2] Or, I say that his answering with this noun ‘man’, as if it desig¬
nated this one species in question—that is to say, one thing, can be attacked generally since
it seems that the utterance ‘man’ is fitted to Socrates, Plato, and all the others in common,
and and so is the name of one thing.

[B] So the words of the text can be read in two ways, either that the answer is given to an
opinion about the thing, or to an opinion about the utterance. (1’) It can be read straight¬
forwardly about the thing in the following way: which, the species, since it is common
to all men, because it is one and indifferent + + + (lac. 8 fere litt.) t is by ment BUT it is
also proper to the incorporeal <species> itself intransitively, that is, is properly this

16) For the introduction of the term materia and the role of P3, see Iwakuma (1996).

17) I am grateful to Y. Iwakuma for letting me use his transcription of P3 (Paris BnF lat. 13368,
215rb-2l6va), forthcoming in the Archives d’histoire doctrinales et litteraires du Moyen Age. See
Iwakuma (1996); Marenbon (2004b). The alphanumeric designations refer to the Catalogue in
Marenbon (2000).

18) See also B10 (Iwakuma 1996, p. 116): “Totum definitivum ut homo constat ex partibus suis,
ex materia essentialiter, ex forma formaliter.”

228

I. Rosier-Catach / Vivarium 45 (2007) 219-237

incorporeal species, because in its purity it is not perceived by the corporeal sense. (2’) It can
be read about the utterance in this way: which, that is the name of the species, although
etc. however it is proper to the incorporeal species, that is, properly signifies this
incorporeal species, as if it were one thing in itself. Thus we say ‘man is a species’ as if it were
a singular proposition. (X) But we should see how, when the utterance ‘man’ is taken in this
way, as if it was the proper name of the incorporeal thing, and afterwards of Socrates and
other men, this does not imply that the name is in some way equivocal—or at least that
the name signifies Socrates by accident, since it properly signifies the incorporeal species.
(Y) Furthermore we should see how we can understand the claim of some people , who say
that this species exists in the divine mind, and is properly signified by ‘man’, and the other
related questions which can be asked about this. 19

The glossator correctly notices the ambiguity of Priscians formulation (text
[2]), and that the passage can be read in re or in voce , 20 a common distinction
in the Glosulae and in the logical commentaries of the time (Iwakuma 1992;
Marenbon 2004b). [A] When one says “What is a rational mortal animal?”,
and someone answers “Man”, and then says that it is “proper to the incorpo¬
real species”, are we questioning and answering about the species or about the
name of the species? Even if, according to the glossator, the passage allows
both a de re reading implying that the claim is about man as the species itself,

19) “Quae quamvis omnium hominum communis esse videatur tamen et propria ipsius
speciei incorporalis. [A] Potest legi et in re et in voce . (1) Posset calumpniari quod dixerat quaeri
de specie quasi de una aliqua re et responded ‘hominem’, quasi unam rem, ideo, inquam, posset
calumpniari /152va/ quod aliquis videret hominem esse in omnibus individuis et ita non posse
<?> unam rem sed plures, et ita posse et debere responderi vel ‘Socratem’ vel ‘Platonem’ etc.
(2) Vel ideo, inquam, simpliciter posset calumpniari, quod responderat ‘hominem’ hoc nomen,
quasi in designatione illius quaesitae speciei unius scilicet rei, cum videret ‘hominem’ hanc vocem
communiter convenire Socrati et Platoni et ceteris omnibus et ideo non esse nomen unius rei.

[B] Et ita duobus modis potest legi haec littera, vel ita scilicet quod respondeat ad opinionem
<de re > vel < ad > opinionem de voce . (1 ’) Secundum rem plane potest legi ita: quae species cum
sit communis omnium hominum quia una et indifferens res + + + (lac. 8 fere litt.) test homini-
bust tamen est etiam propria ipsius <speciei> +++ (lac. 5 fere litt.) incorporalis intransitive,
id est proprie est ipsa species incorporalis, quia corporeo sensu in puritate ilia non percipitur. (2’)
De voce ita: Quae id est cuius speciei nomen quamvis etc. tamen est propria ipsius speciei
incorporalis id est proprie significat illam speciem incorporalem quasi unam rem per se. Unde
dicimus ‘homo est species’ quasi singularem esse propositionem. (X) Sed videndum est quomodo
secundum hoc quod ita accipimus ‘homo’ hanc vocem quasi proprium nomen illius rei incorpo¬
ralis (in corpore ms.) et postea Socratis et ceterorum, non sit quasi aequivocum—vel saltern ex
accidenti significans Socratem, cum proprie significet illam incorpoream speciem. (Y) Deinde
quomodo intellegendum sit quod quidam dicunt illam speciem in divina mente existere et ab
homine proprie significari et cetera quae super hoc possunt quaeri etc.”

20) See Kneepkens (1978), n. 1, which mentions the gloss ‘Licet multi in artes’ where two inter¬
pretations are also offered of the passage.

I. Rosier-Catach / Vivarium 45 (2007) 219-237

229

and a de voce reading implying that the claim is about “homo” as the name of
the species, neither of these readings can confirm Priscian’s claim, and should
be criticized. If (1) the question is about the species as a thing, then the claim
is that the answer “homo” refers to the species as if it was one thing, but since
on this theory man is in all individuals, he is not one thing but many; conse¬
quently the answer would have to be ‘Socrates’ or ‘Plato’ (or the name of some
other individual). If, however, (2) the question is about the name of the spe¬
cies , then the answer would be the name “man”, but this would lead to the
same problem, since the name “man”, given to designate the species as one
thing, , would in fact be the name of many things , i.e. Socrates or Plato. Neither
with the first nor with the second reading can “homo” be the proper name of
the incorporeal species as a singular. Hie second section of the passage [B]
states again the two possible readings of the passage, but with an interpreta¬
tion of “proper” as “properly”. With the in re reading (1’), the question is
about the species man itself, which is common to all men, and which is prop¬
erly the incorporeal species. With the de voce reading (2’), it refers to the name
of the species , meaning that the answer “man” would properly mean the incor¬
poreal species as if it was one thing. Interestingly the glossator, within the in
re (T) exposition, explains further the unity of the species, saying that the
species is common to all men because it is “a thing which is one and indifferent”,
a formulation which is of course reminiscent of William’s second theory of
universal, the one he adopted after the harsh criticisms of Abelard. 21 Unfor¬
tunately the passage is very corrupt here. But maybe it could be interpreted
in the following way. To answer the criticism found in the first section of the
passage [A], grounded on the fact that if the species (taken as a thing or as a
name) was considered simply as one thing being in all individuals, then it
would not be one but many, another solution was devised in [B], that the
species was one as “indifferent”, which thus allows it to be really singular, in
accord with Priscian’s claim. In this indifferentia theory indeed, as known
from Abelard and other sources, the humanity which is in Socrates, Plato,
etc. is not the same one humanity, but a numerically distinct and singular,
although “not different” humanity. 22 It is interesting to find close grammatical

21) Cf. Historia Calamitatum , p. 65: 89-91; Sup. Porph p. 13: 18-14: 6.

22) Cf. De generibus et speciebus , p. 525: “Et sicut de homine dictum est, scilicet quod illud
hominis quod sustinet Socratitem, illud essentialiter non sustinet Platonitatem, ita de animali.
Nam illud animal quod formam humanitatis quae in me sustinet, illud essentialiter alibi non est , sed
illi indifferens est in singulis materiis singulorum individuorum animalis..

230

I. Rosier-Catach / Vivarium 45 (2007) 219-237

discussions in the Sententia de universalibus secundum magistrum R , another
proponent of the indifferentia theory, which should be compared with the
analysis of collective nouns like “ populus \ both in GPma and NDII. 23

The nota ends raising two objections in form of questions. The first one
(X), which is aimed at any interpretation which accepts the dual interpreta¬
tion of “man” as a common and as a proper noun, rightly explains that it
would thus be an equivocal name—an objection already discussed and solved
by the quidam in [5](b), with the distinction between primary and secondary
meaning. The second one (Y), clearly linked to Priscian XVII, 44, wonders, with
some quidam , about how the species can be in the same time in the divine mind
and signified by men—referring back as often in the ND to a previous discus¬
sion which is (unfortunately) not fully reported here.

The gloss on XVII, 43 in the Notae Dunelmenses 1A explains why Priscian
says that even in answers to questions such as “fish swims in the sea”, “fish” is
taken to refer to individuals: Priscian takes into consideration not the value of
the word in this context (as a proper noun for the species), but the actual
action (swimming) which can only be said of individuals, not of a species, a

23) NDII, D, f. 49va, (ad VIII.93; 442.18).

24) ND5, D, f. 157ra-rb (XVII.43, GL3, 134: 21): “Sed videtur hic quoque res individua
esse, etc. Quod dicit in huiusmodi responsionibus ‘bos est utilis aratro’, ‘piscis natat in mari’,
‘bos’ et ‘piscis’ pro individuis accipi, non agit secundum vim vocum et notationes, sed secundum
eventum ipsius actus. Et est simile ei quod excipit hoc verbum quod est ‘tonat’ a tertiis personis
Werborum/ significantibus infinitam personam respiciens, scilicet non ad vim et notationem
vocis, sed ad eventum ipsius actus uni et definitae tantum personae accidentis, id est Iovi secun¬
dum gentiles. Unde et actor, considerans huiusmodi actus ‘arare’, ‘natare’ in solis individuis esse
animati[i]s, dicit (est dicere cod) ‘bos’ et ‘piscis’ in susceptione huiusmodi actuum pro individuis
accipi, non attendens enuntiationem rerum universalium secundum vim et notationem vocum,
id est non attendens dialecticum sensum, sed actum rerum, secundum grammaticum sensum
quern grammaticum sensum etiam in syllabis faciendis et argumentationibus suis attendunt dia-
lectici. Sed quaerendum est quare aratio et natatio non possint esse in re universali, et an similiter
possit dici ubicumque praedicantur quaelibet accidentia, an in quibusdam ita sit, in quibusdam
non, et differentia inter ipsa, et qua ratione discernemus inhaerenti-/157rb/-as rerum pertinen-
tium ad sola individua ab inhaerentiis rerum pertinentium etiam ad universalia, etc. (134.24)
Similiter si dicam multis praesentibus ‘quis est de vobis Grammaticus, quis orator, quis

MEDICUS?’ ET RESPONDEATUR EGO QUIDEM SUM GRAMMATICUS, TU ORATOR, ILLE MEDICUS LOCO

propriorum accipiuntur. Quaeratur an hoc quod determinat multis praesentibus se ita quae-
rere ‘quis de vobis etc.’, id est modus quaerendi aliquam vim faciat ad hoc ut ‘grammaticus’ etc.
loco individuorum in responsione accipiantur, an ipsa responsio ex se sola vim habet secundum
scilicet actum rerum de quibus agit ut in superioribus quoque responsionibus de bove arante et
pisce natante etc. Et de quo sit huiusmodi quaestio ‘quis ex vobis vel quis horum’ etc., an de
inferiori substantia vel alia et qua ratione etc. {fere 4 lin. vacantY

I. Rosier-Catach / Vivarium 45 (2007) 219-237

231

point which could be the object of a further question, the compiler adds. The
nota on XVII, 44 25 merely enumerates, as often is the case in these notae, the
questions which could be raised or have been raised about the passage. It men¬
tions a peculiar claim made by master G., who denied that “ quis” could be
added to an “adjective of substances”, that is adjectives signifying a differentia ,
like “rational”, denying thus the possibility of questions such as “ Quid est
rationale”. The compiler interestingly parallels master G.’s opinion with a
question raised in the “ Topica ” (=Boethius, De differentiis topicis , PL 1187A),
which shows the connection between these passages in Priscian and the Top¬
ics , and the relevance of both for the discussion about the nature of univer¬
sal. Indeed, this very same question is mentioned by Abelard in his own
commentary, shortly after he discussed William of Champeauxs opinion on
predication and inherence. 26 The theological content of the discussion,
already present in question (Y) asked in ND5 (text [7]) is present here with
the strange claim that the explanation of the passage should be made both
“according to God and our faith and opinion (sententia)” and “according to
God and the faith of philosophers and their opinion ( opinio )”, and that both
ways should explain how they understand that divine ideas can be in the
divine mind. This remark is a good witness of the philosophical as well as
theological relevance of Priscian XVII, 44. 27

25) ND5 , D, f. 157ra-rb (XVII.44; 135: 1 = text [4]): “In definitionibus quoque solet etc.
tQuare determinaturf <ad interrogationem omnium> rerum quae definiri possunt et an
agatur de definitionibus tantum substantialibus an etiam de descriptionibus, cum possint et
quaelibet descriptiones dici pro uno substantivo nomini accipi, et qua ratione neget M.G. ‘quis’
adiungi posse adiectivis etiam substantialibus ad interrogationem faciendam, ut ‘ Quid est
rationale etc. : ‘homo’, ut ‘quid est animal ‘substantia animata’. Videtur qui talem dede-
runt animali /157va/ definitionem vel arbori animalia esse voluisse et non animatas. Ex qua
forsitan opinione descendit ilia quaestio in Topicis: ‘Utrum arbores sint animalia necne’ etc.
Quamvis quantum ad generales et speciales formas etc (135: 5). Quomodo continetur
versus iste et quot modis legatur tarn secundum deum (sic) et fidem nostram et sententiam quam
secundum deum et fidem philosophorum et opinionem eorum; et quot modis secundum
utrosque exponi possit ideas esse in divina mente, etc. Et quomodo dicitur in Glosulis in
definitionibus responsis ad ‘quid’ genus et differentias non nominare res suas etc.” We can note
in this passage that master G. and the Glosulae are differentiated, implying once again that the
GPmi are different from Master G.’s teaching on Priscian minor.

26) Abelard, Sup. Top., p. 280: 31, see also 242: 23-24.

27) See the texts in Dijs (1990), p. 95, par. 6 and pp. 114-115, where the same grammatical
analysis of common nouns as proper nouns is found.

232

I. Rosier-Catach / Vivarium 45 (2007) 219-237

Abelard

Priscian XVII, 44 is quoted many times by Abelard, and often along with
views attributed to Plato, in his discussions on universal (Marenbon 1997a,
pp. 186-190). This has led to opposing modern interpretations on the “Pla¬
tonism” of Peter Abelard. 28

As has been fully shown in other studies, the important point in the discus¬
sion over universal is that once Abelard has criticized the view that universal
are things, and chosen the only other alternative that they are utterances
{voces), the question becomes that of the signification of universal names. Since
there is no universal thing to which they correspond, and since Boethius wants
all names and intellections to be grounded in some thing which is its “subject
thing” {res subiecta) , 29 these universal or common names seem empty {Sup.
Porph ., p. 18: 4-19:6; cf. ibid., p. 8: 11-16; p. 30: 6-16). So Abelard, to main¬
tain his position, has to give an answer to three questions: 30 What is the com¬
mon cause of the imposition of common nouns? What kind of common
conception lies behind a universal noun / what is the intellection they gener¬
ate? 31 What justifies that such a name should be said to be “common” (the
common cause of imposition, the common conception, or both)? The answers
involve a long discussion which distinguishes the act of intellection from the
object of intellection, which can be a real thing or a fictive one, and involves a
detailed analysis of the role of imagination and images in the process of cogni¬
tion and signification. Abelard proposes different answers to the problem of
the signification of universal names, which have been interpreted either as an
evolution in his thought, found in the chronological sequence of his writings
(Marenbon 2002, pp. 162-193), or as a conflicting coincidence of various
non-compatible models, occurring sometimes together in the same work,
significantly in the Sup. Porph. (De Libera 1999a, pp. 367-464). I only want
to discuss here one of these analyses according to which common names sig¬
nify mental conceptions, since one of the passages from Priscian is called as the
first authority for it.

28) Tweedale(1976), pp. 185-188; Jolivet (1981); De Libera (1996); Marenbon (1997a), pp. 193-
195; esp. Marenbon (1997b).

29) On the elaboration of this thesis through Alexander of Aphrodisias view of a derived concept ,
see De Libera (1999a), chap. 1 and 2.

30) LI, Sup. Porph., p. 19: 14-20; LNPS, p. 530: 20-23 . See the detailed analysis in Maren¬
bon (1997a) and De Libera (1999a), 367 sq.

31) We find both formulation of the questions, the second one being chosen in LNPS.

I. Rosier-Catach / Vivarium 45 (2007) 219-237

233

In Sup. Porph ., in opposition to Boethius, Abelard interprets the clause that
any noun must have a res subiecta , as meaning that this res subiecta can be
either a real thing or a mental form ( forma concepta ), and that the form can be
either “common” or “proper”, just as to represent a lion, either a general (but
single) picture which represents all lions, or a picture of a lion with character¬
istics which depict one particular lion, can be drawn {Sup. Porph., p. 22:
7-24). The question is then whether the “common form”, that is the one
towards which the mind is directed when hearing a common noun, is the
signification of a universal noun. It is at this point that Priscian s passage is
mentioned (XVII, 44), in which, according to Abelard, Priscian discussed the
imposition of universal names on individuals, also talking about an other
signification they have, of the “common form”. The “general and common”
forms of things mentioned by Priscian, Abelard explains, are exemplary forms,
singular status of things, 32 which are distinguished and known by God before
he produces them by similitude as real things. They are “common conceptions
in Gods mind”, but do not belong to men, who because of their sensible
nature, cannot conceive them as pure forms, but only attached to accidents.
So men know them only in a confused way, as an “opinion” rather than as real
knowledge ( intelligent^ ) (ibid., p. 22: 25-23: 17). The “inventor” imposed
common names on individual things, but also wanted them to signify these
common conceptions, although they are confused and cannot be perfectly
known. For this reason, they are the significates of universal nouns, those to
which the mind is immediately directed when hearing the universal names
(ibid., p. 23: 18-30). Universal nouns are thus common for different reasons,
because they signify a mental conception and because they have a common

32) There are problems to determine whether these common conceptions are to be identified
with status or not, especially when we accept that there are different stages in Abelard’s thinking
about mental conceptions; see the important discussion in Marenbon (1997a), pp. 180-195. In
connection to Priscian XVII, 44, J. M. explains, in a convincing way, that although divine com¬
mon conceptions may be the status of things, it is not the case with human imperfect common
conceptions which are distinct from them (see pp. 194-195). Jolivet, on the other hand, thinks
that the foundation of the signification of universal names is the status existing in God’s mind,
thus seing a kind of platonism in Abelard. I am not sure both explanations are incompatible as
Marenbon puts it (n. 49, p. 194): as a matter of fact, it is possible to accept that the foundation
of universal names is the status as they exist in a perfect state in God’s mind, and as they are
imperfectly thought of (as “opiniones” not as real knowledge) but perfectly aimed at, by the
human mind which is responsible for the actual imposition of words. See also De Libera (1999a),
esp. pp. 373-376 and 461 showing the imprecise nature of the meaning of “status”, which is
taken either for the prototype or for the ectype.

234

I. Rosier-Catach / Vivarium 45 (2007) 219-237

cause of imposition, for instance the noun “man” being imposed for the same
reason on every men (ibid., p. 24: 25-37). Although, as common, they “retain
the similitude of many”, they are also proper since they signify the common
conception (rationality, mortality, etc.) as one (ibid., p. 22: 16-17). This is the
reason for Priscians claim, according to Abelard, that these “common concep¬
tions” are the “quasi” proper names of the universal ( quasi propria nomina esse
dicit ipsa universalia) (ibid., p. 23: 26-37).

But Abelard has more to say about those universal which are names. And
he does so using a distinction, introduced by the Glosulae while commenting
on Priscians definition of the noun as signifying substance with quality ( Insti -
tutiones II, 18; GL2, p. 55: 6-7), between significatio and nominatio. Adopting
a realist position, which is, as we said above, akin to material essence realism,
the Glosulae claim that “man” signifies the quality, a “universal thing common
to all men”, but names individual men. 33 Abelard, having denied the existence
of universal things, says however that “man” signifies a mental conception,
retaining the idea that it names the individual men, adding that it does so
because of the same “cause”, that is because they are all men (ibid., p. 19: 7-
13). Universal names have thus for Abelard a dual signification, a signification
properly speaking, i.e. the generation of an intellection, which is that of the
incorporeal and singular common conception, and a nomination, which is
that of corporeal singular individuals (ibid., p. 29: 35-38). Although the uni¬
versal man was for realists what subsists totally and integrally, at the same
time, in many men, the universal name “man” is, for Abelard, a noun which
names totally and integrally, at the same time, singular men ( tota et integra
singulorum sunt eodem tempore nomina , ibid., p. 31: 11-22). The error of the
realists is then to have improperly transferred a property which is that of names
to things. So, to use Alain de Libera’s words, Abelard showed that what
Boethius had rightly stripped from things, could be legitimately given back to
nouns (De Libera 1999a, p. 488).

To sum up, the common conceptions, which exist eternally in the divine
mind according to Priscian, are at the same time the conceptions men can get
to by abstraction (i.e. considering the thing apart from its accidents), those to
which the mind is directed at when hearing the name. Abelard, in this long
and detailed development seems to give the exact answer the ND compiler was
looking for: how if these ideas exist in God’s mind can they be signified by

33) The passage has been edited and commented on many times, first by De Rijk (1967). See the
edition from all the known mss. of the Glosulae in Rosier-Catach (forthcoming a), annexe 1.

I. Rosier-Catach / Vivarium 45 (2007) 219-237

235

men (cf. Y in text [7]): they are known in their purity by God, but in a con¬
fused way by men.

Abelard gives three authorities to confirm this opinion that universal names
signify common conceptions. The last one is that of Plato. But Abelard denies
that Priscians views are similar to Plato’s, because of the latter’s claim that
universal exist “apart from bodies” ( praeter corpora subsistere ), and because
Plato places these forms not in God’s mind but in noys (ibid., p. 24: 2-24). A
bit later in the text, however, talking about the intelligible and imaginary com¬
position of matter and form which occurs in one’s mind when hearing a com¬
mon noun, Abelard carelessly says that these may be assimilated to Plato s
common and special conceptions (ibid., p. 81: 13-22). In his commentary
on the Peri Hermeneias , Abelard no longer distinguishes Plato’s view from
Priscians, explaining that Plato ascribed the “ideas or exemplar forms” to the
divine mind, and that they are “incorporeal things” {Sup. Per., p. 314: 13-17),
making them the analogue of the imaginary forms in the mind of the crafts¬
man. But in the passage Abelard is talking about the role of imaginary forms,
fictitious images, which are intermediaries, used to talk about real things in
their absence (a view he will develop further in the De intellectibus), not about
the signification of universal names {Sup. Per. p. 315: 30-32). We find the
same assimilation between Priscian and Plato in the later Glossulae super Por-
phyrium (LNPS), when Abelard critically mentions, in the beginning of the
discussion and without further elaboration, the view of those who “think that
universals are intellectus” : for them Priscian seems to imply that mental con¬
ceptions are the universal intellections, whereas Abelard wants to maintain
a distinction between the intellections and their content, or object {LNPS,
p. 513: 15-514: 6). Abelard also uses a short section of Priscian XVII, 44 in his
theological treatises, without any semantic implications, to explain his views
on the divine providence and the procession from the Son, by comparison
with Plato’s doctrine of Ideas as exemplary forms and to Priscians claim that
the generic and specific forms exist in God’s mind before they are effected in
things, which shows that “the concept of the mind becomes operative while it
creates .

Thus it is only in the Sup. Porph. that Abelard accepts Priscians view,
although with a quasi , that universal nouns as proper nouns signify mental
conceptions, but in this text he distinguishes the mental conceptions as they
are in the human mind from the way they exist in the divine mind, and from
Plato’s mental ideas (incorporeal and separated from the bodies). Here, as

34) Theologia summi boni III, 92; Theologia scholarium II, 168; Theologia Christiana IV, 139.

236

I. Rosier-Catach / Vivarium 45 (2007) 219-237

John Marenbon stated, Abelard defends a view which is both non-realist and
non-Platonist. In other texts, as noted above, Abelard will explain the
signification of universal names differently. He seems to move away from the
need to find some kind of real correlate in reality for universal names, as he
stresses more and more the activity of the intellect, which, through its power
of discretio and attentio can “attend to” things in a way different from how they
really are, thus distinguishing even better the intellection as an act from its
content and its object (against the quidam criticized with Priscian in LNPS ),
and elaborating a general solution for all the cases, including the one of uni¬
versal, but also of chimeras, rational stones, laughing asses, propositions, and
the like, where intellections and utterances do not stand in relation of simili -
tudo to things (Rosier-Catach 2004; King forthcoming).

Conclusions

In his article of 1992, Onno Kneepkens suggested that it could not be proved,
as Courtenay (1991) or Normore (1987) had suggested, that “the grammatical
thought of the Priscian tradition was particularly promoting nominalism”, but
also claimed that “grammar turned out to have been neutral to realism and
vocalism or nominalism”. This study suggests a different tentative conclusion:
“the dominating realist atmosphere” of the late 11th and early 12th century—
as Onno Kneepkens calls it—was—at least partially—built on, reinforced by
and transmitted through the reading of Priscian’s Institutiones , and was accord¬
ingly criticized on issues which had to do with linguistic and semantic prob¬
lems raised by grammarians. Moreover, the passage on divine ideas provided a
new source conveying Platonic views on universals, 35 and played a role for
questioning the relation between the mental forms, signified by the common
noun as a proper noun, as abstracted forms (or universal post rerri) and as exist¬
ing in Gods mind (or universal ante rem).

In the Glosulae we found two compatible ways of explaining the signification
of the common noun when used as the proper noun for the species or genus:
it then signifies (1) the universal as abstracted, thought apart from the singular
in which it really subsists; (2) the generic and specific forms, the matter of
things, which are the primeval and exemplary forms in the divine mind. The
core of Abelard s criticism is that these common forms or conceptions do not,

35) It is already found in grammatical glosses connected to Eriugenas commentary on Priscian, see
Luhtala (2000c), and will be found again in the 13th century, see Piche (2003), pp. 194-193.

/. Rosier-Catach / Vivarium 45 (2007) 219-237

237

contrary to what is stated in text [3], exist in rerum natura , but he accepts, in
some passages of Sup. Porph ., that they can be the significates of universal
nouns, explaining in a very detailed way the problem raised in the 7VD, how
the common conception in God’s mind differ from those in man’s mind. He
also answers another question raised by the ND about the equivocal status of
the common noun, if it has two significations. The Glosulae claim that the
signification of the abstracted universal is secondary, and the signification of
the real common universal substance, on which the name has been imposed,
is primary. Abelard borrows from Priscian the idea that the universal name is
imposed on individuals and also signifies a “common conception” {Sup. Porph.
22: 28-30; 29: 37-38), but for him the “common conception” is only thought,
in its purity by God, and in a confused way by men. They all agree that the
common noun names the individuals, but also is the “proper name” of the
universal (which is a mental conception for Abelard and has real existence for
the Glosulae ). The unity of the universal entailed some problems for the gram¬
marians, and we saw that the ND tried two different solutions, corresponding
to the two realist theories held by William of Champeaux.

Further study of the Glosulae and of the ND , in comparison with the testi¬
monies of William on Champeaux’s views, and to the commentaries on the ars
vetus of the early twelfth century, will no doubt shed new light on the major
controversies of the period.

BRILL

Vivarium 45 (2007) 238-252

VIVA

RIUM

www.brill.nl/viv

Symbolism and Linguistic Semantics.
Some Questions (and Confusions) from
Late Antique Neoplatonism up to Eriugena 1

Stefania Bonfiglioli and Costantino Marmo

Universita di Bologna

Abstract

The notion of‘symbol’ in Eriugena’s writing is far from clear. It has an ambiguous seman¬
tic connection with other terms such as ‘signification’, ‘figure’, ‘allegory’, ‘veil’, ‘agalma’,
‘form’, ‘shadow’, ‘mystery’ and so on. This paper aims to explore into the origins of such
a semantic ambiguity, already present in the texts of the pseudo-Dionysian corpus which
Eriugena translated and commented upon. In the probable Neoplatonic sources of this
corpus, the Greek term symbolon shares some aspects of its meaning with other words
inherited from the ancient tradition, such as synthema , eikon, homoiotes. Some of them,
such as eikdn and homoiotes , belong to the field of images and are associated with linguis¬
tic semantics in the Neoplatonic commentaries not only to Plato but also to Aristotle’s
logical works. Among the late ancient Neoplatonists, particular attention is paid to Pro-
clus and to his use of the term agalma. In fact, the textual history of this word seems to be
a privileged perspective from which to reconstruct the Neoplatonic semantic blending of
symbol and image, as well as the main role played by linguistic issues in this conflation.

Keywords

symbol, image, agalma , Neoplatonism, Eriugena

Eriugena on symbola

The translation of the Corpus Dionysiacum and commentary on the Celes¬
tial Hierarchy made by John Scottus Eriugena in the second half of the

0 The responsibility for this paper is equally shared by the two authors. Materially, however,
Stefania Bonfiglioli wrote the second section (‘Neoplatonic agalmata ), Costantino Marmo the
first and third sections (‘Eriugena on symbola and ‘Conclusions). Special thanks are due to Sten
Ebbesen, for his help in all aspects of Stefania Bonfiglioli s work.

DOl: 10.1163/156853407X217740

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

239

ninth century, brought some innovations to medieval semiotics, first of all in
terminology: the word symbolum , which had been previously used in Latin
texts (and by Eriugena himself) as the name for the Creed, is now introduced
as a technical term. 2

Its connection with the word signum is commonly acknowledged by the
theologians of the first half of the ninth century, such as Hrabanus Maurus
and Massentius of Aquileia. They say that the word symbolum means both
index ( indicium ) and collection or covenant (collatio, pactum vel complacitum).
Its second meaning, according to them, derives from the fact that the Fathers
of the Church collected together all the truths of faith in a text, so that these
truths could be heard, known and also memorized by illiterate people. A sym¬
bol, however, is considered also to be a sign (or index) because it is similar to
the military signs that let the warriors of an army distinguish their fellow-
soldiers from their enemies. 3 The Creed was, therefore, considered important
for the identification of a group of people, because of its function of discrimi¬
nating between true and false believers. In Eriugena’s commentary on the
Celestial Hierarchy , the word symbolum keeps its connection with signum , and
in some contexts it refers to a special kind of sign. Symbola , he says, are “signs
that are sometimes similar to sensible things in a pure way, sometimes are not
similar but confused”. 4 Following the pseudo-Dionysius, he explains that
there are two kinds of symbola : those similar to their meanings (such as the sun
as compared to God), and those dissimilar or different from them (such as the
worm as compared to Jesus). We will not spend much time on this distinction,
that is well known and studied. 5 We would like to stress, however, that in
Eriugenas texts there is also a clear semantic connection between terms such

2) A complete semantic history of the term j ymboLon-symbolum has not yet been written. One
can read some chapters of it, for instance, in Ghellinck (1949), 206-207 (with other biblio¬
graphical references to previous works, pp. 273-299), Pepin (1976a), Maestrelli (1976), 793-
795, and Ladner (1979), 223-233. Chydenius (1960) is not very useful, because he fails to take
into account the terms used by the authors themselves.

3) Cf. Hrabanus Maurus, De clericorum institutione , 56, PL 107, col. 396A-B; De universo , 5.13,
PL 111 , col. 136C; De ecclesiastica disciplina , PL 112, coll. 1217C and 1224D-1225A; Mas¬
sentius of Aquileia, Collectanea de antiquis ritibus baptismi , 4, PL 106, col. 55B-D; Epistola ad
Carolum Magnum de significatu baptismi , 4, PL 106, col. 53A-B (both authors depend on the
definition of symbolum given by Isidore of Seville, Etimologiarum sive Originum libri XX, 1906,
§19.57).

4) “Signa sensibilibus rebus similia, aliquando ei pura, aliquando dissimilia et confusa” (Eriu¬
gena, Expositio in Hierarchiam coelestem, 1.2, PL 122, 132C; 1975, 8).

5) See, for instance, Roques (1962, 1967) and Pepin (1976a).

240

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

as symbolum, significatio, typus y allegoria, velamen y visio, imaginatio, agalma y
figura, figuratio y forma, umbra, enigma , and all of them share the property of
being material or sensible. See, for instance, the following text:

He [the pseudo-Dionysius] signified in a variety of ways the above mentioned hierar¬
chies—that are immaterial in nature and above all locations and times, and above all
material shapes, forms and compositions [of forms]—making use of material shapes,
forms and compositions [of forms] imitating our terrestrial hierarchies, i.e. he multiplied
those hierarchies through various visions, symbols, and allegories, taking into account our
[cognitive] weakness. 6

Commenting on the same text, a few pages later, 7 Eriugena puts in opposition
the sensible symbols, through which man can get some knowledge of God,
and the invisible significations (otherwise called intellectual or intelligible
symbols), 8 through which the angels get the same knowledge.

Between symbol and allegory there seems to be a close semantic connection
that might be relevant to clarifying the meaning of the first term, especially if
we take into account that the pseudo-Dionysius does not make use of the latter
word. 9 Eriugena explicitly deals with their relationship in his latest work, his
fragmentary commentary on John’s Gospel, trying to work out a clear-cut dis¬
tinction between symbolum and mysterium. He makes appeal to the distinction
between allegoria in dictis (or in verbis) and allegoria in factis , whose origin
stems back to Augustine but was clearly formulated by Bede. 10 While myste¬
rium is equivalent to both, consisting in “what is handed down according to the
allegory of deeds and of words, i.e. it has to do with historical facts and with
words because [those facts] are told”. 11 As examples he mentions theTaberna-

6) .. predictas ierarchias, id est celestes essentias, dum sint naturaliter immateriales, super
omnia loca et tempora, super omnes materiales figuras et formas et compositiones, per materiales
figuras et formas et compositiones ad similitudinem nostre ierarchie que adhuc in terris est,
uarificauit, hoc est in diuersis uisionibus et symbolis et allegoriis multiplicauit, et nostre
infirmitati conformauit” (1.3, PL 122, 137B; 1973, 13).

7) 1.3, PL 122, 141 A; 1975, 17.

8) VII.2, PL 122, 182A; 1975, 103.

9) Cf. Pepin (1976a) 43.

,0) Cf. Maieru (1999), 139-140.

M) “Mysteria itaque proprie sunt quae iuxta allegoriam et facti et dicti traduntur, hoc est, et
secundum res gestas facta sunt, et dicta quia narrantur... Mysteria itaque sunt quae in utroque
testamento et secundum historiam facta sunt, et secundum litteram narrata” (/« Ev. secundum
Johannem , fr. Ill, PL 122, 344D and 345A; ed. Jeauneau, 352). On this distinction, see also
Jeauneau (1972); Pepin (1973).

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

241

cle, which was built and whose building is recounted in the Scriptures, and the
practice of circumcision, the typical case of sacrament in the Old Testament.
Symbolum , in its proper meaning, is identical with “allegory of words”. The
examples Eriugena refers to are not only the poetical metaphors of the Psalms,
but also the Parables of the New Testament, where Christ tells stories that did
not really happen ( non facta), but were likely to happen {quasi facta)} 2

We would like to underline the fact that this definition of symbol in terms
of allegory does not hold for Eriugena’s commentary on the Celestial Hierar¬
chy. In this text, the case of the Tabernacle (together with all the rites and
sacraments of the Old Testament) is considered as an example of sensible sym¬
bol (whereas the word mysterium , in the same context, is used to refer only to
the sacraments of the New Testament). 13

It is clear that the distinction Eriugena proposes in his latest work does not
apply to his previous ones. According to him, as he has to admit, it does not
even apply to the use of the words ‘symbol’ and ‘mystery’ that one can find in
the Scriptures, where there is no clearcut distinction between them. 14 Further¬
more, the sensible or material character of the symbol that he frequently
stresses in his commentary on the Celestial Hierarchy does not fit (completely)
with the linguistic character of the symbol defined in its proper sense as alle-
goria dicti. The problem is that in his commentary on the Celestial Hierarchy ,
following pseudo-Dionysian suggestions, Eriugena appears to conflate in the
notion of symbol both types of allegory, mixing up metaphor and allegory
into one indistinct concept. Just to add some more confusion, one should
remember that Eriugena makes use of expressions such as “omnes mysticas
figuras, sive in dictis sive in factis”, 15 making clear that the word “allegory” is
probably conceived of as synonym of “[rhetorical] figure”.

Following a path far from the one suggested, for instance, by Pepin (1976a)
who, focusing on the contamination of symbolism by allegory, compared

12) “Altera forma est, quae proprie simboli nomen accepit, et allegoria dicti, non autem facti
appellatur, quoniam in dictis solummodo spirituals doctrinae, non autem in factis sensibilibus
constituitur... Simbola vero, quae solum modo non facta, sed quasi facta sola doctrina dicun-
tur” {In Ev. secundum Johannem , fr. Ill, PL 122, 345A-B; 1996-2003, 352-334).

13) “Sensibilia autem symbola sunt, quorum intellectum caelestes virtutes contemplantur, veteris
legis sacramenta, verbi gratia tabernaculum, et omnia, quae in eo fieri Dominus praecepit,
deinde visiones prophetarum in variis formulis atque figuris, postremo ecclesiastica Novi Testa-
menti mysteria, quorum omnium intimum lumen sancti angeli clare perspiciunt” (Expositio in
Hierarchiam coelestem , VII.2, PL 122, 182A-B; ed. Barbet, 103).

14) Cf. In Ev. secundum Johannem , fr. Ill, PL 122, 345C; 1996-2003, 356.

15) Expositio in Hierarchiam coelestem , II.5, PL 122, 171 A; 1975, 52.

242

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

Eriugenas theories of symbol to the one worked out by the antiqui on allegory
and myth, we will try to uncover some of the Neoplatonic roots of this confu¬
sion, starting from a Greek key-word that occurs in various pseudo-Dionysian
works: the Greek word agalma. As we said above, this word, in Eriugenas
commentary on the Celestial Hierarchy , is tightly linked to the words ‘image 5 ,
‘figure 5 and ‘symbol 5 . 16

Neoplatonic agalmata

Among the few occurrences of the term agalma in Dionysius 5 works, one, in
the De divinis nominibus (IX, 1; PG 3: 909b), is in the plural and specified by
the unusual epithet thednymika. The resulting compound expression, which
sounds almost as awkward as divine-name images, is curious in its association
of linguistic questions with a non-ordinary image, as agalma is. 17 But why did
Dionysius choose just this latter word, so rare in his works, to characterize a
kind of name such as the divine ones? In other words, how, to his mind, did a
name work as an agalma?. Dionysius is not generous with explanations.
Ajiother text, however, may provide us with some clues—a text which may,
indeed, have inspired Dionysius. We are referring to a passage in Proclus’
Commentary on the Cratylus (51: 29ff.; 1908, 19: 8ff.): 18

[The soul], intending to produce similitudes ( homoiotetas ) of the things-that-are—similitudes
that are to be somehow immaterial and derived from logical being only—and availing itself
of the assistance of the linguistic faculty of representation (tei lektikei phantasiai )—creates
by itself the being of names. Just as telestic, by means of certain symbols and mysterious
passwords (dia de tindn symbolon kai aporrheton synthematon) makes the divine images
(< agalmata ) we have here similar (makes similar: apeikazei) to Gods and suitable to receive

16) See for instance the following text, where Eriugena gives some explanation about this word,
which means image, idol or statue (simulacrum), and is very similar to the represented thing or
the model whence it comes: “Agalma autem, ex quo singulari nominatiuo pluraliter agalmata
flectitur, multipliciter etiam ab ipsis Grecis intelligitur. Agalma enim dicunt omnem expressam
imaginem, quae intuentibus letitiam efficit. Agalma quoque dicunt idolum vel simulacrum, ipsi
cuius imago et similitudo est simillimum” (Expositio in Hierarchiam coelestem , II.5, PL 122:
171B; 1975, 52).

17) For the conception of agalma as a particular kind of image, cf. Kerenyi (1962); Criscuolo
(1992), the latter expressly referring to Neoplatonic texts. See Bonfiglioli (2006a, 2006b) for a
history of this notion in the Platonic tradition.

18) For the relation between Proclus’ and Dionysius’ texts, grounded exactly on the occurrences
of the term agalma, see Saffrey (1978).

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

243

divine illuminations, so nomothetic too, in accordance with the same faculty of assimila¬
tion (kata ten auten aphomoidtiken dynamin ), produces names as images ( agalmata) of things
by depicting (< apeikonizomene) the nature of the things-that-are through such-and-such
sounds, and once it has produced names, it gives them to men to use. 19

Proclus here explicitly employs the concept of agalma to elucidate the essence
of names, and the passage is rich in suggesting several connected words that
may somehow clarify the matter. Among those words, the two most recurrent
ones are i) homoiotes , recalled too in the adjective aphomoiotike attributed to
dynamis , and ii) eikon , implied in both verbs apeikazd and apeikonizo. If the
connection here between agalma and eikon is not entirely direct, it is none the
less confirmed by other texts by Proclus, though not expressly devoted to lin¬
guistic matters. Thus, in a passage from his Commentary on Plato’s Republic y
agalmata occurs in a list which includes pictures, shadows and mirror or water
reflections. The feature which agalmata share with the other items lies simply
in their being a kind of eikon . 20

Putting aside for the moment the questions expressly concerning agalmata ,
we should note that the concepts of homoiotes and eikon are also used in other
Neoplatonic texts about the semantics of names. Both notions are employed
by the late-ancient commentators on Aristotle’s Categories when dealing with
Aristotle’s example of equivocation at the very beginning of the work. The
example is zoion, which has two different meanings, ‘living being’ and ‘pic¬
ture’, and can be therefore predicated, respectively, either of a man or of a

19) Cf. the Greek text: “[f] Mmyr \] 3od>.opevt| 8’ avXoxx. ; xporcov xiva mi povr|<; xfj<; Xoyuchg
o\)aia<; eyyovoxx; VTiocrrnaai xcov ovxcov opoioxrixac;, dtp’ eavxfjq, xp&^evr) xfj taKxiKfj tpavxacha
avvepycp, xfjv xcov ovopaxcov rcapfiYayEv ovoiav mi coarcep h xeXeoxucti 8ia 8f| xivcov
<xup(36A,cov mi drcoppf|xcov <Tuv0ripdxcov xa xrjSe aya^paxa xoiq GeoTc; d7ieim£ei mi £ 7 tixf|S£ia
7ioi£i 7tpoq \)7to8oxfi v tcov 0£icov EAAapxj/Ecov, obxco 8 e mi f] vopo0£xiKT) mxa xf]v cruxfiv
cupopoKOXucfjv Suvapiv aya^paxa xcov zcpaypaxcov bcpiaxriai xa ovopaxa 8ia xoicov mi xoicov
hx^v (X7tEiKovi^op£VT| xt^v xcov ovxcov cpuaiv, mi i)7toaxf|CJaaa 7tap£8coK£v Eiq xpfjaiv xoig
dv0p(O7toi<;”.

All translations from Greek are ours.

20) In rempublicam (1899, 1901, I, 289: 21ff.): “[...] xcov 8e a\) eIkovcov SriXobv SvvapEvcov
mi ayaXpaxa mi £(pypa(pfipaxa mi mv oxi xoiobxov, aioxoq [scil: 6 nXdxcov] Siopi^opEvoq,
xivaq Eivai podtaxai xa<; Eimvaq, mi cb<; xaq arco xcov (pcoxi^ovxcov a 7 toxEXo\)p£va<; ev
(pcoxi^opEvou;, xaq xe GKiaq (prioiv EiKovaq mtalv Kai xaq Ep<paa£iq xa q xe ev bSaatv mi xaq
ev xoTq ccXXoiq EvoTixpoiq”.

With regard to the term eikon , whose meaning we shall try to explain in this paper, we would
like to leave it not translated, just choosing to render it, sometimes, by the term ‘copy, in the
contexts where it will be considered in opposition to the term paradeigma .

244

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

drawing . 21 From Porphyry onwards, the brief Aristotelian remarks on this
matter are made to converge into a broad classification of kinds of equivoca¬
tion, on which Sten Ebbesen has already observed the influence of a Hellenis¬
tic theory about concept-formation . 22 Within these discussions, we shall focus
on the choice, in most commentaries, of ranking the Aristotelian example
under the kind of equivocation by similarity . 23 Let us consider, for instance,
Porphyry’s interpretation: according to him, one can use the term zoion both
for a man, qua animate sensitive substance, and for a gegrammenon , qua
homoidma of an animate sensitive substance . 24 There is a first shift from the
source: the gegrammenon which Aristotle had probably intended as picture in
general, without specifying of whom or what, is now related to man consid¬
ered as one of the species of zoion. The word ‘similarity’ is appealed to just in
order to define this added relation between the two equivocal pragmataP But
Porphyry goes on: in another passage, he claims that we call by the term ‘man’
both the living man and the statue ( andrias ) or eikdn. 2(> A further conceptual
shift is here implied, since the equivocal term has become anthropos , and the
words andrias and eikon have been added for qualifying the second pragma.

21) Categoriae (lalff.): “'Optovopa XiyExai tbv ovopa povov koivov, 6 8e mxa xoovopa Xoyoq
xfj<; ouaiaq exepoq, oiov £cpov o xe avBpamoq mi xo yEypappEvov- xofixcov yap ovopa povov
koivov, 6 Se mxa xovvopa "koyoq xrjc; ouaiaq exepoq* Eav yap am8i8co xiq xi Eaxiv auxcov
EKaxEpcp xo £cpcp Eivai, i8iov EmxEpou Aoyov amScoaEi”.

22) Ebbesen (1981), 190-193.

23) Following Porphyry {In Cat. 65: 18-20), this kind of equivocation is not by chance but by
intention {apo dianoias), as well as the kinds by analogy {kat’analogian) , by derivation from the
same source (, aph’henos ), by contribution to the same {pros hen). For the classification of equivo¬
cation kinds in Neoplatonic commentaries, cf. also Luna (1990), 82-100.

24) In Aristotelis Categorias (1887, 66: 22fL): “’E. 'O 8e ApiaxoxE^rjq mice K£xpr|xai xpo7tcp xcov
opcovupwv; A. Ta> am xfjq opoioxrjxoc; Ei7tcc>v £qk>v xov xe avGpamov mi xo yEypappEvov, Eav
yap dm8i8(p xiq xi eoxiv emxepw xo ^cpco Eivai, <Y8iov EmxEpoo A.oyov> omcoScdoei* xov 8 e
mxa xaovopa Xoyov am8i8o\)< ; aXkov aAAcp amScoaEi, xov p£v avGpcomv £(pov Eivai, oxi
ooaia Eaxiv Epxpoxoq aia0r|xiKT|, xo 8 e yEypappEvov £cpov Eivai, oxi opoicopa Eaxiv ooaiac;
Epyoxov aiaGqxiKfjq”.

25) We use the latter expression since both Porphyry and Ammonius, for instance, underline, in
their Commentaries on Categories (respectively, 60: 15fL and 18: 15-16), that Aristotle has
intended to convey the concept of pragmata through the neuter relative pronoun connected with
the adjective “equivocal” (homonyma), in the first line of the Categories (see note 21).

26) In Aristotelis Categorias (1887, 66: 25ffi): “oi 8 e aAAoi am Siavoiaq eIev av fipxqpEvoi, oxi
o xe m0’ opoicoaiv aoviaxapEvoq xpomq xfjq opoovopiat;, oxav <p£p£ dv0po)7tov xe TtpocayopEoaco
xo £a>ov Xoyimv 0 vt|xov mi Eimva avGpcbmo, oxav i8wv Aiya) oxi ‘avGpcomq xooxo’, SqXovoxi
oox ax; Exoxev mi xo ev xq Eirnvi ypappa mA.a> avGpcomv, aAA’ oxi opoicopa Eaxi xoo ^cbvxoq
av0pco7to\) • coco Siavoiaq oov avGpamov xe xov ^covxa Kai xov avSpiavxa q xqv Eimva mA.co
avGpcomv”.

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

245

This pictorial overintepretation may have arisen from the conflation of
different textual influences. Besides the above mentioned Hellenistic theory,
one may also think of some casual remarks by Aristotle to the effect that if
‘man is used of a dead or carved or drawn man, this is an equivocal usage. 27
But the main source is Plato, we think, and his semantics of names in the
Cratylus. In some passages from this work—such as 432b2-d9 28 —names share
with pictures (called in the dialogue either zoia or zdgraphemata) the peculiar¬
ity of being eikones in respect of the pragma to which they refer. The pragma in
question is a man, in this case the man named Cratylus. Moreover, the seman¬
tic connection to their pragma is characterized by the verb aphomoiod , that is
by the concept of similarity.

While Plato’s influence on these matters of logic remains hypothetical,
there can be no doubt, we think, about a genuine Platonic legacy to Proclus’
semantic theories. In the Cratylus passage mentioned above, Plato underlines
that an eikon does not have to be homoion to its pragma in all respects; other¬
wise, if it were, it would result in a double, not an eikon. Hence, the eikon of
the Cratylus is grounded on a semantic selection of its model’s qualities, due to
an act of interpretation. For this reason, in the Sophist Plato links the similarity
of eikones to the concept of heterotes (otherness, difference): an eikon can be
just like its model in some respects but has to be different from it in some
other. 29 This is, in our view, the Platonic basis on which it is possible to under¬
stand the Neoplatonic conception of eikon as involving less informational
content with respect to the paradeigma from which it derives. It is exactly what

27) The same equivocal usage applies to single organs or parts of the body. Cf. especially De
partibus animalium (640b31ff.) and Meteorologica (389b29ff.). For an extempore Aristotelian
classification of equivocation by similarity and analogy, see Physica (249a21fF.).

28) Cf. the Greek text: “Lf2. [...] a-upmoqq eimvoq pq crux aikq <q> q opGoxqc;, vXka to
Evavxiov o\)5e to mpamv 5cq mvxa &7to5ouvai oiov eaxiv cp eim^ei, ei peAAei eIkwv Eivai
[...] ap’ av 5uo 7tpaypaxa ei'q xoiaSe, oiov KpaxiAoq mi KpaxtAoi) eikcov, ei xiq Bewv pq
povov to aov xp^pa mi axqpa d7teima£i£v coa7t£p oi ^coypatpoi, aXka [...] ev'i Xoyw mvxa
(XTtEp cri) £X£i<;> xoiauxa ETEpot mxaaxqaEtEv 7tXqaiov aou; 7tOT£pov Kpaxu^oq av mi eikwv
KpaT'uXov tot’ Eiq to toioutov, q 8i)o Kparutan; KP. Auo EpoiyE 8om\)Oiv, to ScoKpaxEq,
KpaxuXoi [...]. Ifi. TE^oTa yoov, a) KpaxuA.E, vno xcav ovopaxcov m0oi av EKEiva (bv ovopaxa
Eaxiv xa ovopaxa, ei 7tdvxa Ttavxaxq auxoiq opoico0£iq. 5ixxa yap av ttod Ttavxa yEvoixo, mi
o\)K av exoi a\)xcbv eitceiv <o\)5e'k;> ouSexepov 07t6x£p6v egxi xo pev adxo, xo 5e ovopa”. Among
the most recent studies on this passage, see particularly Sedley (2003) 13Iff. For the concept of
eikon in the Cratylus , see also Chiesa (1991), l45ff.

29) Cf. Soph. 234b-24le; 264c-268d. For the conception of image (and of its species) which Plato
draws in the Sophist , see Rosen (1983), l45ff.; de Rijk (1986), ch. 4; Villela-Petit (1991); Carchia
(1997), 52ff; Bonfiglioli (2006a), ch. 2 and (2006b), 30-36, 48-55.

246

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

Proclus claims in the In Cratylum about the names of human language: because
they signify only some features ( semainein ti : 71: 74; 1908, 32: 4) of their
model, they belong at the lowest cognitive rung of Neoplatonist’s hierarchical
ladder.

The term paradeigma too, typically used in Proclus’ texts in connection
with both ontological and cognitive matters, can provide a new link to the
same logical tradition mentioned above, just with regard to the relation already
considered between two equivocal pragmata. Porphyry had defined one of the
two pragmata as an eikon. Ammonius extends this point by appealing to the
relation between an eikon and its paradeigma in order to exemplify the kind of
equivocation by similarity of shape (1895, 22: 8-9) without, however, refer¬
ring explicitly here to the Aristotelian passage. 30 Instead, Philoponus (1898,
17: 8-13), in his first classification of equivocation kinds, and Elias (1900,
140: 21-25) recall the notion of paradeigma! paradigmatic just when dealing
with the equivocal term man’, that is with Porphyry’s revision of the Aristote¬
lian example. In both classifications, however, the kind of equivocation by
similarity has disappeared, and the example just quoted belongs to the kind
aptihenos (see note 23). Following Elias’ Commentary, the equivocation
aph’henos can be defined as ‘from a paradigmatic source’, in the case, for
instance, of a real (alethes) and a drawn ( gegrammenos ) man, for the latter is an
eikon of the former. Philoponus had also added to this ‘paradigmatic source’
the adjective aition —causal, that is. Hence, these passages from logical texts,
in conjunction with Proclus’ ones, show once again that, in Neoplatonists’
view, the connection of an eikdn to its paradeigma —whether this connection
is specified as homoiotes relation or not—must be read in terms of derivation
of an effect from its cause. For this reason, evidently, Philoponus and Elias
have considered the kind of equivocation by similarity as indissociable from
the kind 1 aptihenos

Nevertheless, returning to the relation copy-model in Proclus, other pas¬
sages from his In Cratylum make one realize that a name as an agalma cannot
be easily confined to the description of eikon just offered. For example, in the
lines immediately preceding those quoted above, Proclus argues (51: 19-29;
1908, 18: 27-19: 8):

30) The incipit of the Categories had been already commented by Ammonius in In Cat. 21: 4-7.
3,) This is confirmed by the fact that Philoponus, in his second classification concerning equivo¬
cation (closer to Ajnmonius’one), reinstates the kind by similarity, but considers it just a subspe¬
cies of the kind ‘by derivation from something {In Cat. 22: Iff.).

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

247

Let us briefly tell what the art of creating names is, since the genus of the nomothetic is not
entirely included in it. It is in fact clear that in the soul there is a faculty of producing cop¬
ies ( eikastike dynamis) [...] By means of the same faculty [the soul] by itself makes what is
inferior to it similar (verb: exhomoioi) to itself and even to what is better (ta kreitto) than
itself: for this reason it creates images ( agalmata ) of both gods and daemons. 32

Agalmata are created in order to improve upon the usual copy-model rela¬
tion. Unlike the homoiotes of ordinary eikones , in fact, the similarity of agalma
may become a homoiosis , that means, in our case, a process (marked by the
suffix -sis) of assimilation to the causal paradeigma , a way of tending towards
what is better, which allows one to climb up the cognitive ladder by more than
one rung. Consequently, the list of types of eikones —including agalmata —
that we have already met in the Commentary on Plato’s Republic (see note 20)
has to involve some differences, at least nuances, between the kinds of images
enumerated. This point leads back to Plato, and in particular to the Sophist ,
where image distinctions are grounded on the varying degrees of falsehood,
which any image unavoidably shares in. 33 Not by chance, in fact, the last items
of Proclus’ list—i.e. shadows and mirror reflections—had been already
identified by Iamblichus in the De Mysteriis (II, 10: 20ff.) as phantasmata. With
regard to them, Iamblichus and Proclus add nothing new to Plato’s ranking of
phantasmata as the worst examples of eikon, because of their complete false¬
hood. But something new is introduced by the Neoplatonists when they oppose
agalmata to phantasmata , since agalmata are elsewhere defined as the only kind
of alethes —true— eikon , qua autoptos/autoptikos —i.e. self-revealing. 34

An eikon , in order to become true as only divine things are, must be so
strong as to go beyond its semantic limits of falsehood, that is otherness from

32) Cf. the Greek text: “Tiq 8c tj xwv ovopdxcov tioititucti xexvrj, avviopox; Eirccopev o\) yap mv
caxiv ev a\)xfj xo xr\<; vopo0£xucri<; EiSoq. oxi |iev ovv ecjxi xiq ev \jn)xfj Eimaxuai 86va|j.i<;,
8rAov. [...] mi xa SEVXEpa dtp’ Eavxrjt; E^opoioi [scil: t] xjruxfl] rcpoq eccuxtiv 8ia xfjq a\>xfi<;
8vvd|!£G)<;, mi exi Ttpoq xa KpEixxto Eauxfjq, 816 Begov xe ayd^paxa mi Saipovcov
8tipio\)pyeT”.

33) See note 29. For the comparison between the Sophist and other Platonic works—especially
Republic 10—on these matters, see Halliwell (2002) ch. 1 and 4; Nightingale (2002). For the
passages from the Sophist recalled and interpreted by Proclus in the In rempublicam , cf. Sheppard
(1980), 163ff.; Halliwell (2002), 328.

34) Iamblichus attributes to agalma (in the plural form) the adjective autoptikon in the De Mys¬
teriis II, 10: 23. Proclus employs for agalma the adjective alethes in the In Timaeum (1903-1906,
III, 6: 7), even at the superlative degree; the adjective autopton in the In Timaeum (III, 69: 16)
and in the In rem publicam (II, 133: 17).

248

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

its model, and, consequently, beyond the identity itself as a mere eikon. But by
what means can agalmata be allowed to become like that? “By means of cer¬
tain symbola and secret synthemata\ insofar as they “make them suitable to
receive divine illuminations”: this is the answer Proclus suggests in the first In
Cratylum passage quoted above.

In both Proclus and Iamblichus, the two words symbolon and synthema have
an almost interchangeable meaning, 35 and also often occur together, like in the
passage just considered. According to the different contexts in which the two
authors use them, however, it is possible to point out three distinct ways of
conceiving of symbols.

We shall call the first way too strong. It consists in seeing the peculiar nature
of symbol and synthema to be that they are understood only by Gods and that
they work by themselves without and beyond any human interpretation. 36

The second way is, on the contrary, too weak. According to the second way,
the meaning of the two words is not distinguished from that of signs, or
rather linguistic images, whose creation or explanation depends completely
upon human faculties. It is the case, for instance, of several passages from
Proclus Commentary on the Republic (such as 1903-1906,1, 138: 4ff.), where
the interpretation of the symbols of myths is nothing but an allegorical,
entirely codified, interpretation.

There is also a way in between, which is the Pythagorean one, as reinter¬
preted by Iamblichus and transmitted to Proclus. 37 This way can be exemplified
by an excerpt from Iamblichus’ De vita Pythagorica (1937, 23, 103: 5ffi):

Pythagoras used to pay great attention whenever someone could wisely describe the inner
contents and the mysterious conceptions of the Pythagorean symbols, how much rightness
and truth they share in, once they have been unveiled, freed from their (external) riddling
stamp, and adapted [...] to the natural nobleness of these philosophers—a nobleness made
divine beyond any human imagination [...]. [The members of this school] were not in the
habit of making their words quickly intelligible to listeners [...] but, in accordance with the
silence about divine mysteries enjoined upon them by Pythagoras’ law, they used secret
devices vis-a-vis the uninitiated and covered their mutual exchanges in speech and writing
by means of symbols. Unless one can pick out the symbols themselves and unfold and grasp

35) In agreement with our opinion on the almost interchangeable meaning of the two words, cf.
Sheppard (1980), 146, 152. In disagreement with us, instead, cf. Trouillard (1981), 298-299;
Cardullo (1985). See also about Breton (1981), 317-318.

36) See, for instance, Iamblichus, De Mysteriis (1966), II, 11: 20ff.

37) For the Pythagorean legacy on Iamblichus’ and Proclus’ conception of symbol, cf. Dillon
(1976), 249fF.

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

249

them by means of serious interpretation, their utterances are bound to sound ridiculous
and insane to any listener, and full of nonsense and empty talk. 38

Only the last lines of this passage clarify what Iamblichus means by the sym¬
bolic veils or stamps ( typoi , the latin figurae) covering the inner core of the
utterances of the Pythagoreans. 39 Through such figurative expressions Iambli¬
chus conveys the need for interpreting symbols on two different levels. The
first level, the serious one according to the Neoplatonic author, is reached only
by the few sharing the elitist code of the Pythagorean school. The second one
is accessible to the majority of people, but cannot go beyond a more superficial
interpretation of the utterances, unable as it is to recognize them as symbola. It
follows that symbolic questions are not concerned here with a distinct kind of
signs named symbola. Instead, they deal with a way of conceiving certain signs,
whether eikones or utterances or so on, qua symbola , that is as liable to be inter¬
preted according to two different codes.

Just this Pythagorean legacy may have motivated the Neoplatonic associa¬
tion of the words symbolon and synthema. Before its occurrences in Neopla¬
tonic texts, synthema is in fact a word belonging in particular to the
historiographical tradition, where it means “password”, especially military
password. As such, it is a form which had currency within a narrow (military)
circle and its elitist interpretation of certain signs.

Recognizing a sign as a symbol does not, however, imply disconnecting its
meaning from the conventionalism shared by any interpretative code which
can be varied, taught and learned by any number of people. 40 The difference

38) Cf. the Greek text: “[...] mpd riuGayopa pEyd^qg cnto\)8fjg ETuyxavEV, ei Tig SiapGptoaEiE
oacpcbg Tag xtbv n-uGayoptKwv auppotaov EpcpaoEig mi d7toppr)Tou<; Evvotag, oaqg opGoxrjTog
mi dXr|0£iag petexovoiv d7tomA.u(p0Eiaai mi xob aiviypaxcoSoug EA.£u0£p(o0Eiaai xxmou,
7tpoaoiK£UD0£iaai 8e [...] xaig xcov (piA^oaocpcov tovtcdv p£yataxp\uaig mi UTtEp dv0pa)7uvr|v
£7uvoiav GecdGeioi. [...] o\) auvExa etcoiouvxo [scil. : oi ek tou 8iSaam^£iov xouxou]
ETtiSpoprjg xoig aKovovat [...] xa cppa^opEva [...], aXXa mxa xf|v v£vopo0Exr|pEvr|v auxoig
\)7to flvGayopou ExepvGiav Geiouv puaxripitov mi rcpog xoug axeXecTovq a7toppT|xcov xpo7tcov
t^tixovxo mi 8ia auppotaov £7t£aK£7iov Tag 7tpog aXXr\Xovq SiaXE^Eig q croyypacpag. mi ei pq
xig auxa xa crupPoXa EK^E^ag StaTtxu^EiE Kai apcoKtp E^qyqaEi <7t£piA.dPoi>, yE^oia av mi
ypatoSq So^eie xoTg Evxuyxdvovai xa tayopEva, A.'npo'u pEaxa Kai aSotaoxiag”.

39) The Iamblichean opposition between superficial appearences and hidden contents of the
symbols is recalled and developed by Proclus especially in the In rem publicam , once more in
comparison with the way of signifying by means of eikones. For a study on the passages involved,
which are too many to be quoted and examined in our paper, cf. particularly Coulter (1976),
39ff. See also Pepin (1976a), 56ff.; Trouillard (1981), 299-302.

40) After all, according to Neoplatonic commentators, the very signification by convention ( thesei)

250

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

between the two steps of interpretation is rather located in what a password
allows only few to do. Especially in the case of circles of initiated, the knowl¬
edge of an elitist symbolic meaning, though still clinging to the limits of a
rational interpretation, is after all the access to another experience far beyond
these limits, that is to an ascent towards the divine. It is on this same opposi¬
tion between Limited and Unlimited that in other texts Iamblichus bases the
comparison between symbols and seeds of nature. 41 Both of them look
superficially very easy to deal with but, from both, an unforeseeable multiplic¬
ity of effects as well as of hidden contents can emerge. In virtue of this link to
infinity, symbols seem to meet again the first, too strong way of being con¬
ceived, that is the mystical way related to the divine, which makes any human
analysis fail.

However, even confining the interpretative ascent to rational steps, it is
possible to conclude that an agalma is at the same time an eikon and a symbol,
being an eikon symbolically interpreted in the sense just described. In Proclus
and then in Dionysius, divine names can thus be distinguished from simple
nam cs-eikones, in virtue of the further meanings they can receive, obviously
only if recognized as symbola.

Following this linguistic path of nam es-agalmata may also provide a privi¬
leged look into the evolution of the more general conception of symbolism
from late Antiquity to the early Middle Ages. Let us come back to the Diony¬
sian corpus , now considering an excerpt from the De ecclesiastica hierarchia
(I, 2; PG 3: 373 a-b):

We see our [hierarchy] multiplied {plethynomenen ) in proportion to ourselves by the variety
of the sensible symbols ( tei ton aistheton symboLon poikiliai ) by which we—in the measure
proper to us—are led up ( anagometha ) hierarchically to the uniform deification. They,
being intelligences, understand in accordance with what has been permitted them, whereas
we are led up C anagometha ) to the divine contemplations by means of sensible ( aisthetais )
eikones, to the extent that this is possible. 42

characterizes the linguistic symbols of Aristotle’s De Interpretations (l6al-8, 27-28; 24bl-2). See
Proclus’ In Cratylum (47: 1-3; Pasquali 15: 27-29) and Ammonius’ In Interpretationem
(20: 1-8; 23: 5-9, 30-33).

40 See, for instance, De vita Pythagorica 29, 161: 5-10.

42) Cf. the Greek text: “ttiv [tepapxiav] m0’ fipaq 8e opojpev ava^oyax; qpiv avtoiq xf\ xcov
aia0r|T(ov avppoXcov TtoiKiXia 7t>.T)0\)vopivT|v, ucp’ dbv iepapxiKccx; aii rfiv evosi5ri Oetooiv ev
auppetpia xr\ Ka0’ dvayope0a. Ai pev <bq voeq voodai Kata to avtaiq 0epit6v, fipeTq 8e
aio0nta!q eiKoatv ejtl taq 0eia<; ax; 8\)vatov dvayope0a 0ea)piaq”.

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

251

In Dionysius, symbola share with eikones the adjective ‘sensible’ ( aistheta) and
the link to words such as ‘variety’ ( poikilia ) and ‘multiplicity’ {plethos , recalled
by the verb plethynd in the passage just quoted). They are thus characterized in
terms which Iamblichus and Proclus had reserved for eikones , i.e. for the signs
related to the lowest step of human knowledge. In Dionysius, then, eikones
share with symbols the verb anagein , which Iamblichus and Proclus had
reserved for symbolic ascent. Consequently, from Dionysius onwards, the
conception of symbols is subject to an even greater confusion between sym¬
bols proper and eikones , and in fact it becomes impossible to keep the two
notions apart. It is no coincidence, we think, that it was the same kinds of
signs that Proclus, while avoiding Dionysian terminological blending, had
already fused into agalmata , especially in the linguistic context of the Com¬
mentary on the Cratylus.

Conclusions

The examination of some probable sources of the pseudo-Dionysian corpus
leads us to some provisional conclusions.

First of all, both images and symbols are characterized, in the pseudo-
Dionysian works, by their materiality or sensible quality, and both share an
anagogical function according to Eriugena: these features place the Eriugenian
symbols at the first level of the materialis manuductio (which translates the
Greek phrase hylaia cheiragogia) that can lead believers to the knowledge of
God. The word agalma , because of its meaning of “divine-name image”, seems
to play a pivotal role towards the blending of so many different types of signs
and ways of signification.

Secondly, the excerpt from Iamblichus’ De vita Pythagorica has shown that
the Pythagorean conception of symbol, in general, suggests a connection
between terms such as symbolon, typos (stamp, but also figure) and parapetasma
(veil). The influence of Neoplatonic sources seems particularly relevant in
some texts where Eriugena stresses the need to go beyond the sensible features
of sacred symbols and figures (also in the sacrament) to grasp divine truth: he
addresses his criticism not only against those who are only interested in a lit¬
eral interpretation of the Scriptures, but also against some “new” theologians,
such as Paschasius Radbertus, who held that the Eucharistic bread and wine,
that hide the physical presence of Christ’s body and blood, exemplify a par¬
ticular kind of signs which are identical with their meanings. Paschasius’ inter¬
pretation of the Eucharistic rite was aimed at making understandable (and

252

S. Bonfiglioli, C. Marmo / Vivarium 45 (2007) 238-252

almost perceptible) the mysteries of the incarnation and of the Eucharistic
presence to a large number of mainly illiterate people—and it also used to this
end a dossier of Eucharistic miracles. By contrast, in line with his Neoplatonic
sources, Eriugena stresses the esoteric function of symbols and sacraments,
whose meaning only a few learned persons can attain. 43

Finally, the meaning of synthema as a military password goes in the same
direction. With reference to this feature, we would like to stress the unex¬
pected link between the commonly adopted definition otsymbolum (as Creed)
we started with above and the concept of synthema : this suggests another path
to follow through the Patristic sources of the Synods where the Creed was
defined as symbolon.

43) On the ninth-century Eucharistic debate, see Marmo (2003) and (2005) for further biblio¬
graphical references.

BRILL

Vivarium 45 (2007) 253-268

VIVA

RIUM

www.brill.nl/viv

“Utrum idem sint dicere et intelligere sive videre
in mente”: Robert Kilwardby, Quaestiones
in librum primum Sententiarum x

Mary Sirridge

Louisiana State University

Abstract

In his Questions /, qq. 35-36 Sent. Robert Kilwardby asks whether divine understand¬
ing ( intelligere ) is the same as the divine speaking {dicere), as Anselm says in Monolo-
gion , ch. 63, just as for us mental speaking {mentis locutio) is the same as the thinkers
examination {inspectio cogitantis) or mental seeing {videre in mente). His answer is that
neither for us nor for God is the equation correct, because understanding lacks an
essential characteristic of speech, i.e. referentiality, and because speaking is active and
understanding passive, which is reflected in the meanings {impositions) and gram¬
matical functions {modi significandi) of the corresponding expressions. Kilwardby does
concede in his discussion of the speech of angels in II Sent. q. 56 that when inner
speech does occur, and remains internal, it amounts to thought, though with the addi¬
tional element of referentiality. I suggest that Kilwardby is unwilling accept Augustine’s
theory that thought is inner speech, as Anselm does, because it would require him to
reject Aristotelian-style philosophical psychology.

Keywords

Anselm, speech of angels, divine Word, modes of signifying

In his Questions on Book I of the Sentences I, 2 Robert Kilwardby asks whether
divine understanding {intelligere) is the same as the divine speaking {dicere).

0 Research for this paper was generously supported by an A.C. Mellon grant from the Knights
of Columbus Vatican Film Library of St. Louis University. I follow the orthography of the pub¬
lished text of Kilwardbys commentary on the Sentences , and of Schmauss edition: S. Anselmi
Cantuariensis Archiepiscopi Opera Omnia , even at the cost of inconsistencies of capitalization
with respect to divine names.

2) Kilwardby (1968).

DOI: 10.1163/156853407X217759

254

M. Sirridge / Vivarium 45 (2007) 253-268

as Anselm says in Monologion, ch. 63; following Anselms wording more
precisely, 3 he asks whether for God it is the same to speak ( dicere ) as to “gaze
upon something by thinking” ( cogitando intueri ), just as for us mental speak¬
ing ( mentis locutio) is the same as the thinkers examination ( inspectio cogitantis)
or mental seeing ( videre in mente) (I Sent., q. 36, 372-8). The identification
of speaking and understanding seems improbable, he points out, even if there
can be no real distinction in being or operation in God, since speaking and
understanding are different by nature and definition. Kilwardbys resolution
(I Sent., q. 36, 423-30) is that if we attend to what is properly meant by
speaking ( loqui) and understanding {intelligere), to speak {dicere) is not the
same as to gaze upon something by thinking {cogitando intueri) or to understand
{intelligere).

Kilwardbys answer is unexpected for a number of reasons, not the least of
which is that he seems elsewhere in this same work to accept the characteriza¬
tion of thought as a kind of inner speaking. In the preceding question, speak¬
ing of the Trinity (I Sent., q. 35), he argues:

In a mind that is always contemplating itself there are are always these three: the begetter,
the begotten, and the mutual love joining them... This second premise is proved according
to what Augustine says, cnamely that> thought ( cogitatio ) is a word born from the knowl¬
edge ( notitia ) in memory... Augustine teaches that thought is a word born of the knowl¬
edge ( notitia ) in memory in De Trinitate XV, c. 24 [=De trin. 15.x. 19] as follows:
“Necessarily when we speak of what we know, there is born from the knowledge ( notitia ) 4
we have in memory a word that is of exactly the same sort as the knowledge {scientia) from
which it is born. And thus the thought ( cogitatio) formed from the thing we know ( scimus )
is a word that we say in the heart.” You must understand that Augustine is not speaking
here about the spoken word or spoken language, but about the mental word. Anselm says
the same in Monologion , ch. 48: “The word of the thing is the thought itself formed in the
likeness of the thing from memory” (I Sent., q. 35, 126-42).

3) Anselm (1968), 73.

4) The wording of De trin. XV.x. 19 is “ex ipsa scientia quam memoria tenemus.” Though ‘notitia
occurs in De trin. XV, it is the preferred terminology of the verbum-xhtory of De trin. IX. The
basic terminology of Kilwardbys own discussion is: videre, intelligere, cogitareI cogitatio and dicere
or loqui ; the last two expressions he seems to use interchangeably, often giving both conjoined
by sive. Frequently, at least, he uses cogitare to indicate the process of thinking, and intelligere for
the simple grasping of a concept; cogitatio is used indifferently for thinking, having a thought,
and the thought itself. Anselms formulation from Monologion, ch. 62 adds the relatively
undefined and nontechnical inspectio and intueri to the mix; Kilwardby himself seems to use
intueri consistently for cases in which the mind searches for or finds something within itself
(cf. II Sent, q. 41; I Sent., q. 62, 118-123).

M. Sirridge / Vivarium 45 (2007) 253-268

255

And in his commentary on the second book of the Sentences , 5 6 discussing the
speech of angels (locutio spirituals angelorum) he appears to accept the argument
that goes: since the human spiritual word ( verbum spirituale) is thought,
as Augustine says, the speech of angels must likewise just be their thought
(II. Sent., q. 56, 7-10).

As these examples indicate, the thought of Augustine casts a long philo¬
sophical and theological shadow over Kilwardby s entire commentary on the
Sentences', he is clearly familiar with Augustine’s verbum- theory of thought and
understanding, and with the identification of inner thought, inner vision and
thought or understanding that finds its most explicit expression in De Trini-
tate XV. He has appropriated Augustine’s distinction between external words
and the verbum cordis, between language as an external system of signifiers and
an internal domain of the signified from which external language takes its
meaning and its capacity for denotating things. Whatever the Stoics them¬
selves meant by distinguishing spoken and written language from lekta, G
Augustine’s acute consciousness of the distinction between a system of external
signifiers and what they signify internally looks back to this Stoic distinction,
even if the characterization of the signified as internal language is a Porphyrian
twist. 7 One obvious source of such ideas was the work of the stoicizing gram¬
marians and rhetoricians with whom Augustine was familiar; whatever their
exact intellectual antecedents, these theorists consistently assumed the exis¬
tence of a complex structure of meanings that underlies the (more or less adept
and grammatical) utterances of spoken and written language. 8

Kilwardby himself would also have drawn a generally Stoic orientation to
meaning and semantics directly from the stoicizing grammarian Priscian
by way of his own work as a grammarian. Priscian Minor (Priscian, 1859)
announces its subject as the construction of the well-formed or “perfect” state¬
ment ( oratio perfecta) (IG .XVII. 3); in a famous passage, it traces this gram¬
matical well-formedness to the meaning of the words ( intellectus vocis) intended
by speakers and understood by hearers (IG XVII. 187). In his commentary on

5) Kilwardby (1992).

6) A.C. Lloyd (1971), 63-66, defends the view that Stoic lekta were the theoretical equivalent of
Aristotle’s concepts; according to A.A. Long (1971), ontologically speaking lekta are placed
somewhere between the merely mental and the external realities to which language refers. For a
similar explanation, see also Ebbesen (2004).

7) Cf. Ebbesen (2005a) for a general discussion of basic Stoic themes and their permutations in
the Middle Ages.

8) For a discussion of Stoic semantics and grammar, cf. Ebbesen (2005a), 68-70; Colish (1985),
184-189; Long (1971).

256

M. Sirridge / Vivarium 45 (2007) 253-268

Priscian Minor , Kilwardby adopts the view, common among medieval gram¬
marians, that such meanings can be conveyed from person to person because
there is an established inventory of vocabulary assignments and combinatorial
possibilities. 9 Even deliberately deviant constructions, the various figurae , are
to be explained in terms of exploiting this inventory (Sirridge 1990).

It is, of course, a considerable jump from a stoicizing theory about linguis¬
tic meaning to psychological verbalism, i.e., from holding that complete and
grammatical statements will be just those that have a certain kind of structural
correspondence with complete thoughts to the view that thinking and thoughts
are essentially linguistic. Augustine, however, made just this leap in De Trini-
tate\ at least this is surely how Anselm understood Augustine’s position. If
Kilwardby is nosing out from under Augustine’s long shadow, then some other
element in the mix of his original or inherited ideas is outweighing Augustine’s
influence. There are, I think, two such elements. First, the Stoic-inspired the¬
ory that thought is inner speech has come up against the predominantly pas¬
sive picture of sensation, memory and cognition favored by Aristotelian-style
philosophy of mind; Kilwardby is understandably loath to depart from the
best scientific orthodoxy of his time. Secondly, it is Kilwardby s view that we
cannot say that verbs of speaking and verbs of knowing mean the same thing,
strictly speaking, because to do so erodes their meanings and grammatical
functions; the farther expressions get from their proper meanings, the greater
the danger that they will become meaningless altogether.

Eighty-One Sons of God and Still Counting

It having been established in the previous question that there are at least three
divine persons, Question 36 asks: Whether there can be more than three? The
structure of Question 36 is extremely complex. The question begins with four
lengthy positive arguments that there can be at most three divine persons.
These arguments are fairly predictable: Suppose there is a fourth person; since
the divine persons differ only with respect to origin, that person must be dis¬
tinguished either as God that comes from God or God from whom God
comes; if it is God from whom God comes, there is already one of these, from
whom the supposed fourth person is indistinguishable, etc. There is a series
of eight main objections to these opening arguments. In the course of the
second of these objections, the one Kilwardby says has to do specifically with

9) Cf. Ebbesen (1983), 68-73; Ebbesen (2004); Sirridge (1988, 1993).

M. Sirridge / Vivarium 45 (2007) 253-268

257

the identification of the Son as Word, we find an argument ending with the
perverse conclusion that there are eighty-one—or perhaps an infinite number—
of sons of God, and as many Holy Spirits:

But it seems that there are here many Words, and thus many Sons. For each of the three
speaks itself and the other two. But there are as many words as there are things said, and for
any given thing spoken about there are as many words spoken about the thing as there are
speakers. Likewise 10 if several men think the same thing there seem to be as many words of
the heart as there are thinkers; and if anyone thinks several things, there are in his heart as
many words as there are things thought. But Anselm maintains in Monologion 62, where he
treats this question, that each of the three says itself and the other two. And the same con¬
clusion follows from his words in c. 35, i.e., that “for the highest Spirit speaking is the same
as understanding.” Again, c. 32: “For Him, to speak and to understand are the same.” And
further on in the same chapter: “If it knows itself eternally, then it says itself eternally.” And
so, just as each of the three knows itself and the other two, so it speaks itself and the other
two. And so there seem to be at least nine Words, but really an infinite number if you like.
This is because each of the nine says itself and the others, just as before; and whoever speaks
does so with a word, because whoever says a word also says with a word whatever he says.
Thus there will be nine times nine Words, and as a result as many Sons, and so on to
infinity (I Sent., q. 36, 224-39).

By the way, he adds, there will also be an infinite number of Holy Spirits, for
one will proceed from each Father and Son pair!

It is within the context of this this objection that Kilwardby argues that
speaking and understanding are demonstrably not the same, either for us or
for God. Without the crucial identification of inner speaking and thinking,
the fact that each member of the Trinity thinks itself and the others does not
lead to a proliferation of divine Words.

Kilwardbys discussion of whether understanding is identical with inner
speaking is complicated by the fact that Anselm’s Monologion is involved here,
as it was in the discussion of the previous objection. 11 In this second objection,

l0) Etjquia MW +Schne ider. Monologion , which Kilwardby is following, has placed after ‘quot
sunt cogitantes: ‘quia in singularum cogitationibus verbum eius est’ {Mon. 62, 72).
n) The conclusion of Kilwardbys first objection runs: “Item ex hoc videtur quod Pater sit ali-
cuius filius et alicuius spiritus, et quod Filius sit alicuius pater et alicuius spiritus, et similiter de
Spiritu Sancto quod sit alicuius pater et alicuius filius” (I Sent ., q. 36, 215-217). As he notes,
this question is nearly identical to the question Anselm poses in Monologion 61 (Anselm 1968,
71). Kilwardby adds a second undesirable result of Anselms claim: “ibi sit trina memoria, trina
intelligentia et trina dilectio, et sic trinus Pater, trinus Filius, trinus Spiritus Sanctus” (I Sent.,
q. 36, 213-215. This argument is not explicit in the Monologion 61 (Anselm 1968, 71), though
Anselm concludes his answer to the question with, “Quid igitur prohibet concludi quia unus
tantum est in summa essentia pater, unus filius, unus spiritus, et non tres patres aut filii aut
spiritus?” {Mon. 61, 72).

258

M. Sirridge / Vivarium 45 (2007) 253-268

the reasoning of the Monologion has again been used in setting up the objec¬
tion: If, as Anselm says in the Monologion , each of the three persons under¬
stands itself and the others, and if for God it is the same to speak as to
understand, then if a divine Word is the Son, we must ask why there are not
nine Sons of God—and indeed there will actually be nine-times-nine Sons of
God if each of the nine thinks itself and the other eight, and so on to infinity.
Anselm himself posed this question, Kilwardby adds. 12 Then, as in the case of
the first objection, Kilwardby offers an answer that is a paraphrase of Anselms
own answer from Monologion c. 63 [Mon. 182-4): Since divine knowing or
understanding is not done by each person individually, though each thinks or
understands, there is only one divine knowing and one divine understanding,
not three; and since the divine understanding is identical with divine speak¬
ing, just as our thinking is identical with our mental speech, there is also only
one divine speaking; and since the divine speaking speaks the divine nature,
the divine speaking really speaks only one divine Word.

As before, the Anselmian answer gives rise to further objections: Even if the
divine activities are in reality identical, 13 nonetheless if we go by the meaning
of the words used ( intentiones verborum) and how we understand them ( ratio -
nem intelligendi) , Anselm is wrong to suppose as he clearly does, that divine
speaking is identical with divine understanding.

[1] To speak is to point to or refer to something, but understanding is a kind of vision. Making
a comparison with seeing, speaking has to do with the thing already seen, and understand¬
ing with what is going on in the person doing the seeing.

[2] If we take the isomorphism with bodily activities at all seriously, then like external
speaking, inner speaking has the character of an action; and hearing is the kind of being
acted upon that corresponds to speaking. If inner hearing and inner seeing are the same, as
Augustine says they are in De Trinitate (=XV.x.l8), and if understanding is inner seeing,
then understanding will be inner hearing. Thus it seems that understanding and inner speak¬
ing are opposites by definition; and if they are, then Anselm is wong to say that say that in the
supreme Spirit, this kind of speaking is nothing but grasping by thinking (cogitando intueri ),
just as the speech in our minds is nothing but the gaze of the the apprehension of the thinker
(inspectio cogitantis) (I Sent., q. 36, 372-386.).

Kilwardby s answer is brief and to the point. Anselm is wrong, if he means what
he says literally. For if we are speaking and thinking strictly and correctly, inner

12) Anselms conclusion in Monologion 62 (Anselm 1968, 73) is less elaborate than Kilwardbys,
but the intent is the same: “Hac itaque ratione videntur in ilia esse non solum multi patres et filii
et procedentes, sed et aliae necessitudines” (Mon. 62, 73).

,3) idem \om. WM+Schneider.

M. Sirridge / Vivarium 45 (2007) 253-268

259

speaking and understanding or “grasping by the thinking” are not the same.
Nevertheless the two activities do run together towards one real act of grasping
or knowledge (ad unam actualem contemplationem vel cogitationem concurrunt),
and somehow they flow together to single end, as if from opposing directions
(quasi ex adverso) Perhaps this is what Anselm meant, Kilwardby adds; or per¬
haps he meant to refer to the divine unity of essence (I Sent., q. 36, 425-30).

The Inner Word and the Speech of Angels

Matters have not been settled in Question 36. The reasoning of Anselm has
been dismissed, and with it the threat of eighty-one or more divine Words.
But the dismissal comes at a price, for if the Fathers speaking forth of the
Word cannot be glossed in terms of his self-understanding, it becomes unclear
what the Father, and only the Father is doing when he begets the Word, and
only the Word and why the Word should be called a word at all. These prob¬
lems are taken up in subsequent objections (I Sent., q. 36, 394-411). Kilwardby
addresses the problems by appealing to distinctions between speaking in the
strict and proper sense (proprie) and speaking loosely or improperly (commu-
niter sive improprie) (I. Sent., q. 36, 431-95). It is in fact difficult to see how
this distinction helps to distinguish the speaking proper to the Father from a
speaking more loosely common to all three divine persons. But if our interest
is mainly in Kilwardby s relationship to his Augustinian-Stoic source material,
there is a more pressing question: If Kilwardby rejects the identification of
inner speech and thinking or understanding, then what is to be made of the
passages in which he seems to accept Augustine’s verbum-xheory?

What are we to make of Kilwardby’s discussion in I Sent. q. 35 of how the
Father generates the Son, where Kilwardby has quoted with apparent approval
Augustine’s identification of inner vision and inner speech from De Trinitate
XV.x.10. In. Q. 35 he makes no attempt to gloss or reinterpret Augustine’s
words, perhaps because nothing inconvenient follows upon the identification
of inner speech and thinking or understanding. But in I Sent. q. 36, where the
identification is called sharply into question, the reading Kilwardby has tenta¬
tively assigned to Anselm’s position is offered explicitly as a gloss on what
Augustine says in De Trinitate XY: “Speech is one thing and vision another
when they are done outwardly through the body, but inwardly, when we
think, both are one (Aliud est locutio et aliud visio, quando foris per corpus
fiunt, intus autem cum cogitamus, utrumque unum est)” (De trin. XV.x .10).
What Augustine meant to say in De Trinitate XV, it is proposed, was that
mental speaking and vision “run together into a single knowledge and are one

260

M. Sirridge / Vivarium 45 (2007) 253-268

in this sense” (in unam coeunt cogitationem , et sic sunt unum) (I. Sent. q. 36,

388-393).

We have not the time to make a complete assessment of Kilwardby’s reading
of De Trinitate XV; but we can make a start by asking what Kilwardby means
by this somewhat mysterious claim that inner speech and inner vision or
knowing “combine forces” or “flow together” or “come together as if from
opposite directions to bring about a single knowledge or thought.” And here,
it seems, we are in luck, for in Question 35, in the course of demonstrating
that there are at least three divine persons, Kilwardby offers a fairly detailed
account of the workings of the mind, ostensibly directed to showing how we
should understand the generation of the Word by the Father.

The account starts from a quotation from De Trinitate XV that “when we
speak, that is when we speak what we know, from that knowledge that we hold
in memory there is born a word which is exactly of the same sort as the knowl¬
edge from which it is born. Indeed, the thought formed from the thing which
we know is a word that we speak in our heart.” (De trin. XV.x. 19). In corporeal
things, Kilwardby continues, true motion or action occurs when a natural
agent introduces into matter or draws out of the potentiality of matter a form
similar to its own; we have generation when the natural action of an agent
results in a single composite of matter and form. The process of perception is
similar, though here there is no real action, as Aristotle says in De Anima III;
here the sensible as agent (sensibile agens) impresses its likeness on the sense
power, bringing about an alteration, so that the sense power in act (sentiens in
actu ) is what is generated by the sensible species that makes the impression.
What happens in the thinking or knowing mind is exactly parallel. The “gaze
of the mind” (odes mentis) that is not yet thinking has the power to know or
understand, and in memory there is lodged the knowable thing itself (rem
cognoscibilem ) or some likeness of it that is potentially knowable or thinkable.
When the “gaze of the mind” that is thinking or understanding turns to the
image in memory, that image remains in memory, just as the object of sense
remains while it is being sensed. And so

... memory conjoined with the gaze of the mind makes that gaze like itself by impressing
upon it the likeness it has within itself (not by sending forth from itself the likeness it has
in it, but by impressing one similar to its own upon the gaze of the mind.) So there is the
likeness in memory that impresses and the likeness impressed upon the gaze of the mind,
and a single offspring is made from the impressed likeness and the gaze of the mind; this is
understanding {intelligentia). In sensing the sensible form remains in its matter after its
likeness has disappeared from sense upon the sense turning aside to sense another object,
and the sense power in act is once again born from this sensible form when the sense power

M. Sirridge / Vivarium 45 (2007) 253-268

261

turns back to it; just so, in the intellect the intelligible species remains in memory after the
likeness that it impressed on the gaze of the mind has perished from the mind’s gaze because
the gaze of the mind turns towards thinking something (cogitandum) else, and then once
again understanding is born from it when the gaze of the mind turns back towards it
(I Sent., q.35, 145... 176).

Kilwardby seems to think he has given a flat and incontestable account of the
verbum- theory of De Trinitate. And indeed, some key elements from De Trin -
it ate are in evidence here, most notably the “gaze of the mind.” We also find
Augustine’s notion of memory as a broadly representative power. As Kilwardby
notes elsewhere, this power differs from the Aristotelian understanding of
memory as a storehouse of past sense images in two important respects: there
is memory of things known, not just of things sensed; and memory functions
as a representative power even in occurrent perception and thought. 14 What
we do not find in Kilwardby’s account, however, is a connection to an Augus-
tinian verbum-i\\toty. In this key passage, we do not find an attempt to con¬
nect the explanation of how understanding is begotten or formed from
memory with the genesis of some inner word. At no point does Kilwardby
announce firmly: And here there is begotten a word. 15 It is significant, I think,
that the Augustinian terminology that Kilwardby fills into his causal account
is drawn in the main from Augustine’s account in De Trinitate Xl.iii-ix and
XIV.iii,vi, where Augustine is occupied with presenting the visual model of
cognition, and not from IX.vii-ix and XV.x-xi, where he presents the verbal
model.

In fact, in substance, the account Kilwardby gives here is closer to De
Anima than to De Trinitate. The shift to generatio/genitum terminology allows
Kilwardby to substitute an Aristotelian-style account of motion, change,

14) In I Sent. , q. 62, 50-105, Kilwardby gives a full discussion of the Augustinian notion of
memory, which includes both “memoria brutalis,” which we share with lower animals, and
“memoria rationalis.” Rational memory is again divided into a lower sort of rational memory,
which records the “imagines rerum corporalium per sensus adquisitas de quibus est visio spiri¬
tuals sive imaginativa,” and a higher memory of spiritual, incorporeal things (spiritualia de
quibus est visio intellectuals.) Rational memory both precedes understanding and retains things
that have been understood. Cf. also I. Sent., q. 59, 102-111.

15) That he means to defer this announcement until the next section on the “mutual love copu¬
lating begetter and begotten” is unlikely, as this later section is obviously already intended as an
illustration of procession, which has to do with the of the holy spirit. And as Kilwardby would
know, at De trin. IX.x.15, Augustine distinguishes between two kinds of inner word, one of
which is present whenever something is known or thought of.

262

M. Sirridge / Vivarium 45 (2007) 253-268

sensation, and cognition; there is, for example, no trace here of Augustine’s
consistently held view that the mind sends the senses out for information
(De trin. IX.iii3: XI.iv.7), or of his view that the concepts of the mind, many
of which precede empirical experience, inform the data of sensation from the
start (De trin. X.v.7). More generally, despite some Augustinian talk about
“aversion of the sense” and “conversion of the mind to thinking something
else,” lifted more or less directly from De Trinitate Xl.iii-ix, for Kilwardby
sensation and cognition are not active powers, as they are for Augustine. For
Kilwardby, sense sensing and the mind thinking or knowing are, of course, in
act; but they are expressly designated as “what is generated.” They are thus at
most re-active, because they are put into act by the species expressed and
impressed upon them by the object of sense and the object of thought lodged
in memory, respectively. Only the roving “acies mentis ” has an Augustinian
sound, though this terminology really comes from the visio-xhzory of De
Trinitate XIV.

Kilwardby seems to think that he has given a way of reading Augustine that
is essentially compatible with Aristotelian psychology. Arguably, he has not,
though his knowledge of Augustine’s thought is extensive and subtle. 16 For
Augustine, the idea that thought or understanding is an inner word had an
integral part to play in a theory of mind, divine and human, that was essen¬
tially active; presenting understanding as formation of an inner word was
Augustine’s first step in establishing thinking as inner speech. 17 The verbum-
theory does not play a corresponding role in Kilwardby s psychology. Kilwardby
gives us no account of how the inner word is formed or of what makes it a
word; and so his interpretive claim that Augustine means to say that under¬
standing and inner speaking “come together as if from different directions to
bring about a single knowledge or thought” remains a mere promisory note.

Kilwardby’s discussion of the speech (locutio) of angels in his commentary
on the second book of the Sentences (II. Sent., q. 56), is another instance of his

l6) Kilwardby’s “Tabulae,” in which he summarizes book by book and explains some thirty-one
works of Augustine, probably postdates his commentary on the Sentences (Callus 1948, 255-
263). But his commentary on the Sentences gives evidence of a good deal more acquaintance with
Augustine’s works than the knowledge that Stone (2001) assigns to most thirteenth-century
figures: “a few standard treatises... and abundant quotations that circulated under his name”
(Stone, 2001,255-256).

,7) Sirridge (1999) argues that Augustine did intend in De trin. XV to argue for the identification
of thought and inner language, and not for the less extreme theory Kilwardby ascribes to him
that inner speech amounts to thought, and that inner speech and inner vision (or thought or
understanding) combine to produce a complete actual act of contemplation or understanding.

M. Sirridge / Vivarium 45 (2007) 253-268

263

seeming to accept the identification of thought and inner speech. Here the
larger context is the question of what happens when angels are “sent” to make
announcements and the like. The particular question concerns the speech of
angels, not when they speak “by the bodies they assume, but their spiritual
speech” (II. Sent., q. 56, 5-6). It is argued that such angelic speech must
amount to angelic thought {cogitatio) —because angels resemble us with
respect to their spiritual intellects, and a fortiori with respect to their spiritual
words, so that their spiritual words amount to thoughts, just as ours do, as
Augustine says in De TrinitateYN (II. Sent., q. 56, 5-10). It is then objected
that if angels’ thoughts were speech, since speech serves essentially to make
determinations of one’s will manifest to others, angelic thoughts would imme¬
diately be broadcast every which way, so that lower angels would know every¬
thing the higher ones know (II. Sent., q. 56, 16-19).

To this counterargument, Kilwardby responds with a distinction between
(i) spiritual speaking that is just thought, i.e., speaking absolutely, in oneself
and contained within one self; and (ii) spiritual speaking directed to another
and thus made manifest, in which something is thus added to thought itself. 18
God’s thoughts are always in a sense contained within himself; even when He
speaks to a soul or an angel, He thinks what He thinks ( suum cogitare ), and
does not undergo change—and accordingly ( juxta quod) at a certain point in
the temporal order Jonah apprehends that God is telling him to go on a mis¬
sion. Creatures sometimes just talk to themselves in the absolute sense, within
themselves; and then their spiritual speech is just thought. But they always
speak to each other in the second way, so that on the side of the speaker some¬
thing is added to the thought, namely expressing or manifesting it ( superaddit
locutio super cogitationem ); on the side of the hearer there is also an effect,
namely that the hearer’s mind apprehends in one way or another ( impressio
aliqua in audiente) what the speaker wants it to apprehend.

It looks initially as if Kilwardby is just equivocating on dicere/loqui by say¬
ing that inner speech is the same as thinking, as Augustine says, except when
it is more than thinking, i.e., when the thinking is communicated. He is, I
think, doing a bit more than this. He is first trying to explain how angels’spiritual
speech can amount to thought without every thought of every angel being
known to every other angel. His explanation is that sometimes angels just talk

18) II. Sent., q. 56, 23-25: “Solutio. Est loqui dupliciter, vel in se et absolute apud ipsum dicen-
tem et haec est cogitatio, vel relatum ad alterum secundum quod est manifestatum et sic super¬
addit aliquid super cogitationem.”

264

M. Sirridge / Vivarium 45 (2007) 253-268

to themselves, just as we talk to ourselves, “contained in ourselves.” Kilwardby
is on firm enough phenomenological ground in saying that for us such “spiri¬
tual speech” is nothing more than thought. It is hard to see why things should
be different for angels, since, as Kilwardby sees it, angels have matter and indi¬
vidual Augustinian memory (though no senses), in which they have represen¬
tations of individuals located in space-time, just as we do. 19 But sometimes we
talk to others; and given our embodiment, this will involve something more,
namely making our thoughts manifest to them in some corporeal way. In their
spiritual speech to other creatures, angels are essentially similar to us; they, too,
have to add an element of directing and manifesting their thoughts, though
this will not involve physical speaking. 20 And thus, although their spiritual
speech to others is thought, it is thought with direction or relationality added
to it. Kilwardbys undertaking is ambitious; this account is intended also to
explain how God can speak to His creatures, e.g., how he can send Jonah on a
mission or send the Angel Gabriel to speak to Mary, which happens at a par¬
ticular time, when in God there is only unchanging thought. Here the answer
seems to be that some of God’s thoughts, too, have a direction. They have this
direction eternally, and God thinks his thoughts about Jonah and his mission
eternally, and God’s thought is eternally directed to Jonah-at-a-time. And at
that point in time a creature like Jonah or Gabriel gets the impression of being
spoken to by God.

But if this is what Kilwardby is up to in II. Sent., q. 56, then the issues
under discussion here are different from those of I. Sent , qq. 35-36, though

19) Kilwardby comes close to accepting the knowability of singulars in his earlier De Ortu Scien-
tiarum (Kilwardby 1976); cf. Sirridge (1988), 15-18. II. Sent., q. 37, 94-131, “De scientia dae-
monum naturali et angelorum ,” shows that he now fully accepts knowledge of singulars qua
individuals, very likely because he now has fully at his disposal Augustine’s distinction between
higher and lower kinds of knowledge.

20) Kilwardbys view resembles that of Aquinas to some extent. In order to speak spiritually to
humans and other angels, angels have to direct or “utter” their thoughts to their hearers. But if
Aquinas has angels communicating by angelic telephone (Panaccio, 1999, 325), Kilwardby has
them sending graphic images to each other via their videophones; their images are like ours,
except that their images of things are directly caused by God, and not by the things of which they
are images. But unlike Aquinas (Panaccio, 1999, 230), Kilwardby does not hold that angelic
minds are transparent, for they have spiritual matter. Because of their great “subtlety” (subtilita-
tem) and simplicity, they can make themselves immediately present to other minds and penetrate
into the inmost depths of another thinker (this is how demons find out what people are thinking
as well.) With other angels, then, they can impress their images on the other angel’s mind, as they
do with humans, or they might just somehow direct the other angel’s attention to those of its
own infused species the first angel wants it to think about (II. Sent., q. 42, 5-11).

M. Sirridge / Vivarium 45 (2007) 253-268

265

not completely unrelated to them. Kilwardby is not saying here that all thought
is inner speech, let alone that they are the same thing, but rather that when
inner speech stays within, it amounts to thought. Direction to another mind
or the intention to manifest a thought is something added to thought that is
already inner speech in the absolute sense. The intention to manifest a thought
is thus inessential to its being spiritual speech. The mode of the manifestation
is even further removed from the essential. We have here an excellent example
of one of Kilwardby’s favorite strategies: “speaking” in itself or taken abso¬
lutely has the same meaning for God, angels and humans, the relevant
differences being dependent on how certain relations, themselves not essential
to speaking, are realized.

This position on inner speech is consistent with what Kilwardby says in his
commentary on the first book of the Sentences. The issue at that point was
whether thinking ( cogitare ) or understanding ( intelligere ) is exactly the same as
inner speech. It is significant that understanding ( intelligere ) is not in play in
the discussion of the speech of angels in II. Sent. q. 56, where Kilwardby reso¬
lutely uses cogitatio!'cogitare. This discussion thus leaves open the possibility
that understanding is inner vision, and therefore is not inner speech at all, and
for that matter that some thought is not inner speech. Indeed the two discus¬
sions are to some extent complementary. II. Sent., q. 56 does not tell us what
characteristics are essential to that kind of thought which does amount to
inner speech. I Sent., q. 36 supplies this information by implication; the very
characteristics of speech that led Kilwardby to conclude that inner vision or
understanding cannot be understood as inner speech pick out the thoughts
that are inner speech. First, such thoughts refer to or point to things under¬
stood; they are referential. Second, thought that is inner speech is active (in
contrast to inner vision or understanding, which is passive.)

Conclusion: Imposition and Modes of Signifying

Despite being replete with quotations from Augustine, with respect to the
“inner word,” Kilwardby’s commentary on the Sentences does not present us
with an unreserved assimilation of Augustinian philosophical psychology, for
Kilwardby is rejecting the thesis that thought can be understood as mental
language, though he accepts the weaker claim that when there is genuine inner
speech, it amounts to thought. Kilwardby is clearly interested in using Augus¬
tinian philosophy of mind to complement Aristotelian-style psychology, rather
than to challenge or replace it.

266

M. Sirridge / Vivarium 45 (2007) 253-268

The role of the Monologion in this discussion of inner speech is extremely
significant. In his commentary on Book III of the Sentences , 21 Kilwardby
describes Anselms thought in Cur Deus Homo? as “probably right and subtle”
(III Sent., q . 4, 41-2). But the Monologion is frequently criticized in connec¬
tion with the alleged identification of understanding with inner speaking.
Indeed, it seems to be introduced, so as to have a clear target for arguments
against a thesis that Kilwardby considers wrong. Kilwardby also means to
make the point that if the Monologion is taken literally, its elegant madness is
a misrepresentation of the thought of Augustine.

What is specifically wrong with the Monologion , it seems, is its use of words
like “understanding”, “speaking” and “seeing” in violation of the normal sense
for which they were established. This reading is confirmed by looking at
Question 72 of I. Sent.: “Whether the Father and Son love each other by the
Holy Spirit?” Here we get further information about the difference of “mean¬
ing of the words and in how we understand them” between verbs of speaking
and verbs of knowing. This time the issue of speaking and understanding
comes up in the course of an argument that divine loving and divine being are
not the same:

We can agree that being [esse) and exercising wisdom (sapere) are the same for God, not just
because of identity in the thing signified, but also because of the modes of signifying, for
“esse” and “ sapere” both signify absolutely and in the mode of being-at-rest (in quiete) (cor¬
respondingly, opposing modes of signifying will mean that what <the expressions> say is
not the same.) In contrast, “ diligere” signifies transitively and in the mode of being-in-
action (in fieri), and these are opposing modes 22 of signifying. And so it should not be
agreed that for God to be and to love are the same. Moreover, the verb of speaking or say¬
ing, having to do with cognition, is closer to being an absolute verb like “esse” than the verb
of loving, which has to do with feeling. And so if divine speaking or saying is not the same
as divine being, divine loving is even less so (I Sent. q. 72, 299-307).

This argument, Kilwardby says, looks like a good one, but it runs into difficulty
if, as Anselm says in Monologion , understanding is nothing other than inner
speaking. In this question, Kilwardby does not explain the resulting “difficulty”
further, and contents himself with numerous quotations to show that Anselm
does indeed identify divine speaking and divine understanding or knowing in
the Monologion. But his idea is probably that if Anselm is right, then our “good”

21) Kilwardby (1982).

22) modi]modis W+Schneider; modi modo M.

M. Sirridge / Vivarium 45 (2007) 253-268

267

argument can be turned around: If divine speaking and divine knowing—or
the exercise of divine wisdom—are the same, then, since divine knowing and
divine being are the same, divine speaking and divine being will be the same by
transitivity of identity. But then if divine speaking and being are the same,
despite the somewhat different modes of signifying of verbs of speaking and
“esse”, then modes of signifying cannot be all that important, and so divine
loving and divine being may also be the same, despite somewhat more different
modes of signifying of “diligere” and “esse”!

But Anselms view that divine speaking and divine knowing are identical is
not right, Kilwardby rejoins, not at any rate unless Anselm means that in the
divinity there is no real multiplicity, or that understanding and inner speaking
“combine to produce a single knowledge or contemplation of the mind
{concurrunt ad unam cogitationem vel mentis contemplationem ejficiendam)”
(I. Sent., q. 71, 499-503). (Probably that is what he means, Kilwardby adds.)
But nonetheless:

If we give any weight to the established meanings of the expressions (intentionibus) it will
be obvious that speaking or saying is somehow a cause of understanding, and that under¬
standing comes from speaking. Thus for the reason we gave previously, they are not the
same, and also because there is a different mode of designating and mode of understanding
{designandi et intelligendi) for speaking or saying and for understanding. This is because the
act of understanding terminates in the mind, as it were, and occurs in the quiet of contem¬
plation, while the act of speaking or saying comes forth from the mind, as it were, and is
carried on in doing, and in action or motion. For this reason it does not follow that any
ablative whatsoever that truly qualifies (< determinate ) one of these verbs can also truly
qualify the other. And therefore even if someone says or speaks by means of a word, it can¬
not truly be said that he understands by means of a word. Taking the expressions in this
way, speaking or saying is not the same as understanding (I Sent., q. 72, 503-13).

Thus Kilwardby is rejecting the approach of the Monologion because it distorts
and erodes the impositiones and modi significandi of words by which their
semantic and syntactic capacities are fixed so that a conventional external lan¬
guage can be used to communicate what the speaker intends to the under¬
standing of the hearer. The alternative to observing the impositiones of words
is the evaporation of meaning. Oddly enough, then, in the case of Robert
Kilwardby s approach to the question of whether inner speaking is the same as
inner seeing or understanding or thought, Aristotelian psychology and a sto-
icizing approach to linguistic meaning square off against a Stoic-inspired the¬
ory about the nature of rational thought—and they prevail.

With respect to this important issue, there is more juxtaposition than
assimilation in Kilwardbys thought. Historically speaking the juxtaposition

268

M. Sirridge / Vivarium 45 (2007) 253-268

may be as significant a phenomenon, for it is reasonably clear that confronta¬
tion with Aristotelian theory has produced the promise of a re-reading of
Augustine’s theory of inner speech. As we have seen, Kilwardby does not really
make good on his own promise of a non-Anselmian account of Augustine’s
theory, but it is significant that he thinks there is one. We have a promise, that
is, that the two theories may “run together as if from different directions to
create a single thought or knowledge.”

BRILL

Vivarium 45 (2007) 269-282

VIVA

RIUM

www.brill.nl/viv

Mental Language and Tradition Encounters in
Medieval Philosophy: Anselm, Albert and Ockham

Claude Panaccio

University of Quebec at Montreal

Abstract

Medieval philosophy is often presented as the outcome of a large scale encounter
between the Christian tradition and the Greek philosophical one. This picture, how¬
ever, inappropriately tends to leave out the active role played by the medieval authors
themselves and their institutional contexts. The theme of the mental language provides
us with an interesting case study in such matters. The paper first introduces a few
technical notions—‘theme’, ‘tradition, ‘textual chain’ and ‘textual borrowing’—, and
then focuses on precise passages about mental language from Anselm of Canterbury,
Albert the Great and William of Ockham. All three authors in effect identify some
relevant Augustinian idea (that of‘mental word’, most saliently) with some traditional
philosophical one (such as that of‘concept’ or that of 'logos endiathetos). But the gist
of the operation widely varies along the line and the tradition encounter is staged in
each case with specific goals and interests in view. The use of ancient authoritative texts
with respect to mental language is thus shown to be radically transformed from the
eleventh to the fourteenth century.

Keywords

compositionality, mental language, theme, tradition, verbum

In the very first chapter of his Summa logicae , William of Ockham introduces
his celebrated view of the concept as the basic unit of the language of thought.
To warrant the notion, he refers in so doing to two of the most outstanding
authorities of the medieval intellectual world: Boethius and Augustine, bor¬
rowing from the former the idea of a mental sentence (oratio mentalis) and
from the latter that of a mental word belonging to no conventional language
0 verbum mentis)? This is a typical case where the old Greek philosophical

0 William of Ockham (1974), 7.
© Koninklijke Brill NV, Leiden, 2007

DOI: 10.1163/156853407X217768

270

C. Panaccio / Vivarium 45 (2007) 269-282

tradition—represented by Boethius—and the Christian religious tradition—
personified by Augustine—directly meet with one another within the limits of
a single text, even a single paragraph in this particular instance. Medieval phi¬
losophy as a whole, indeed, has often been seen as the outcome, on a larger
scale, of precisely such encounters. But how exactly does that work? Are we to
suppose that traditions copulate somehow and give birth to offsprings? An
image of this sort is sometimes assumed in the historiography of philosophy.
Yet it must certainly be spurious, if we think of it, in so far as it leaves no active
role to play in the end to the medieval thinkers themselves! Whatever they are,
traditions after all, are not agents.

The medieval theme of mental language handily provides us with the material
for an interesting case study in such matters. My approach in the present paper
will be to scrutinize three short, but significant, passages from different authors—
Anselm, Albert the Great and William of Ockham, namely—where the idea of
mental language, or some version of it, is explained with the help of precise, if
sometimes indirect, references to both Greek philosophy and the Augustinian
corpus. I will build in the process upon the results of my previous synthetic study
of mental language from Plato to Ockham, 2 but the analysis will be conducted
in a fresh perspective, with a special concern for how intellectual traditions typi¬
cally interact in medieval philosophy. My aim is to show how, with respect to
mental language, the use of ancient texts and ideas is completely transformed
from Anselm to Ockham, the reason for it being that the theoretical preoccupa¬
tions radically changed in the meanwhile, mainly because of new scientific and
institutional developments. Textual traditions will appear, in this case at least, as
much less causally productive by themselves than is often imagined.

Traditions and Themes

Before getting on to the selected passages, however, a few words are in order
about the general outlook I will be adopting and the main operational con¬
cepts I will be using. History of philosophy is frequently conducted as if the
basic actors of intellectual developments were impersonal entities such as tra¬
ditions, currents of thought, doctrines, schools, religions and so on. This
might be handy most of the time, but it should never be forgotten that those
are mere convenient abstractions. The individual thinkers and their particular
spoken and written utterances are, ultimately, the only realities out of which

2) Panaccio (1999).

C. Panaccio / Vivarium 45 (2007) 269-282

271

traditions and the like are made up. The history of philosophy is but a scat¬
tered plurality of utterance events connected with each other in various ways.
The idea of a “theme”, such as the theme of mental language, and that of a
“tradition”, such as the Aristotelian or Augustinian traditions, are simply
unintelligible unless understood in connection with concrete and spatio-
temporally located human activities.

Let us start with the notion of a theme , which is an important component
of the approach to be pursued here. The relevance of comparing the three texts
I have chosen is grounded, prima facie , in the fact that they are supposed to
share a given theme, the theme of mental language namely. Yet such a “com¬
mon theme” is not to be thought of as a transhistorical entity of its own that
would be the real extra-textual referent of the various texts under consider¬
ation. As I have explained elsewhere, a theme in the history of philosophy
should be seen as something very similar to a musical theme? It is not located
anywhere out of the texts themselves, just as the musical theme is not located
anywhere out of the musical sequences. The only realities in the latter case are
the concrete musical performances. To say that two such performances exhibit
a common theme is but to say that they resemble each order in repeatedly
displaying similar strings of notes. In the same vein, two different philosophi¬
cal texts can be seen as having a theme in common if they crucially display
similar strings of words.

The similarity of words in such cases, however, should not be thought of
as merely phonological, but as primarily semantic. Two texts built up from
widely different phonological units—belonging to different languages, for
instance—can legitimately be said to exemplify the same theme if a certain
string of words that is centrally present in one of them has some close seman¬
tic equivalence with a string of words centrally present in the other one. In
the particular case of mental language, my habit has been to say that a certain
text, from whatever period, exhibits the theme of mental language when it
crucially contains certain typical phrases composed of at least two parts, one
of which normally refers to the order of mental phenomena while the other
normally refers to the order of linguistic phenomena, such as “ oratio mentalis ”
and “ verbum mentis ’ in Latin, “ logos endiathetos ” in Greek, or ‘mental language
and “internal discourse” in English. Some extra-linguistic reference is assumed
here, of course (how could there be semantic similarity without it?), but it need
not be to a special abstract and trans-temporal entity. And, most importantly,
it need not even be the global reference of the complex phrases at issue: only the

3) Panaccio (2001), 262-265.

272

C. Panaccio / Vivarium 45 (2007) 269-282

usual reference of their simpler components to concrete mental or linguistic
phenomena is relevant for the theme in question to be present.

As to traditions , they are best construed as groups of concrete textual chains.
What I call a textual chain in this context is a chronologically ordered series of
texts with a common theme, such that each text in the series (except the first)
is linked with the previous one by a direct borrowing in connection with the
common theme, a borrowing, here, being any form of integration of some
parts or other of the previous text into the new one. A literal quotation would
be a salient example, of course, but several other varieties of borrowing are
possible, including oblique reports, adoption of typical phrases, distinctions
or arguments, and outright plagiarism. A borrowing from a text A by a text B
will be said to be “direct” only if the reading (or hearing) of text A actually
played a causal role in the use of the borrowed string by the author of text B ,
and it will be said to be indirect if some intermediate text is required between
A and B to account for the borrowing.

The idea of textual tradition can then be introduced on the basis of textual
chains. This is not the occasion to engage in a detailed discussion of the best
way to do it. Various non-equivalent notions should be introduced, obviously,
to take into account the various interesting ways in which textual chains can
be grouped for historiographical purposes, and a much more complex—and
more precise—set of definitions could thus be constructed along these lines.
But let us simply agree at this point that what we will call a textual tradition
in the present paper (or sometimes a tradition tout court) will be any large
cluster of textual chains rooted in a single well delimited corpus such as the
writings of Aristotle or Augustine or the Fathers of the Church and so on.

However informal and incomplete, these indicative explanations should at
least make it apparent that the causal links which constitute a scattered plural¬
ity of texts as a chain, and a plurality of textual chains as a tradition, necessar¬
ily require the intellectual activity of the later authors: reading and writing in
particular. The relevant causality, in the end, is merely a form of mental causa¬
tion, and it always runs, in reality, from a particular author (or group of
authors) to another one by the mediation of externalized written or spoken
utterances. The presumption, then, is that the intentions, preoccupations and
interests of the borrowers, the institutional and social constraints they are sub¬
jected to, the audience they want to reach, and in general the particularities of
the context in which they are reading and writing, are crucially determining
factors as to what is borrowed exactly, how the borrowing is done, and what
import the borrowed strings acquire within the new context. This, emphati¬
cally, should always be kept in mind whenever we engage in intellectual history.

C. Panaccio / Vivarium 45 (2007) 269-282

273

Traditions and themes are all too easily reified into autonomous entities of their
own—a way of proceeding which frequently leads to mythological narratives
rather than sound historical work. Tradition encounters, in particular, when
they occur in definite texts, should be analysed not as the mere results of
“influences” or the mere recurrence of previous ideas, but as active elaborations
by given authors in given contexts. Let us now turn to specific instances.

Anselm of Canterbury: Monologion , 10

Our first sample is from chapter 10 of Anselm’s Monologion. Having argued in
the opening chapters that there is one and only one Creator, Anselm then
reflects, starting with chapter 8, upon the notion of a creation ex nihilo. In
chapter 10, specifically, he introduces the idea of mental speech (mentis locutio
in this case) in view of comparing the creative thought of God with the inter¬
nal discourse of the craftsman planning to realize something. Here is the text
(or at least the parts of it I will be interested in here):

But what is this form of things, which in the Maker’s reason preceded the things to be cre¬
ated, other than an expression-of-things [rerum locutio] in the Makers reason?—just as
when a craftsman who is about to make a work from his craft first speaks of it within him¬
self by a mental conception? Now, by ‘mental expression’ [locutio mentis] or ‘rational expres¬
sion’ [locutio rationis] I do not mean here thinking the words which are significative of
things; I mean, rather, viewing mentally, with the acute gaze of thought, the things them¬
selves which already exist or are going to exist. For in ordinary usage we recognize that we
can speak of a single object in three ways. For we speak of objects either (1) by perceptibly
employing perceptible signs (i.e., [signs] which can be perceived by the bodily senses) or
(2) by imperceptibly thinking to ourselves these same signs, which are perceptible outside us,
or (3) neither by perceptibly nor by imperceptibly employing these signs, but by inwardly
and mentally speaking of the objects themselves—in accordance with their variety—either
through the imagination of material things or through rational discernment [...]

Each of these three kinds of speaking has its corresponding kind of words [verba]. Yet,
words of that [kind of] speaking which I mentioned third, and last, are natural and are the
same for all races, if they are not words for unknown things [...] [N]o other word seems so
similar to the object for which it is a word, and [no other word] so expresses that object, as
does that likeness [similitudo] which is expressed in the acute gaze of the mind as it con¬
ceives the object itself. Therefore, the natural word is rightly to be called the principal and
most proper word for an object. 4

4) Anselm of Canterbury (1946), 24-23.

274

C. Panaccio / Vivarium 45 (2007) 269-282

The passage obviously exhibits our theme of mental language since the phrase
“locutio mentis ’ or “ locutio rationis ” occurs in it in a central role: it is indeed
this kind of internal speech that Gods creative thought is supposed to be illu-
minatingly compared with. No author, ancient or recent, is named or quoted
in the text, but two sources are easily identifiable: Augustine’s De Trinitate and
Aristotle’s Peri Hermeneias .

From Augustine comes the distinction between three ways in which we can
speak of something. Anselm’s enumeration, admittedly, is not found in this
exact form in the De Trinitate , but it comes very close to Augustine’s division
of the different senses of “verbum” in book IX, chapter 9, paragraph 15; 5
and, most importantly, the important opposition in Anselm’s text between

(2) silently processing within ourselves the conventional linguistic signs, and

(3) inwardly thinking the things themselves without the help of conventional
signs, obviously corresponds to the distinction repeatedly stressed by Augus¬
tine in book XV between the mental word properly speaking, which belongs
to no conventional language, and the mental representation of linguistic units
(e.g. De Trinitate XV, 10, 19). It is very doubtful that Anselm, who, we know,
read the De Trinitate with utmost care, reinvented these ideas independently
and they can be considered, therefore, as direct borrowings, albeit non-literal
ones. Anselm’s passage, moreover, also relies—at the end of the first paragraph,
when it further distinguishes between two ways of mentally speaking of the
objects themselves ‘either through the imagination of material things or
through rational discernment”—upon Augustine introducing first a merely
imagined word of the thing itself in book VIII (e.g. 6, 9) before turning in the
later parts of the De Trinitate to a more spiritual kind of mental word. And,
finally, the reference to the mental viewing of the things themselves “with the
acute gaze of thought” ( acie cogitantis) toward the beginning of Anselm’s text
is, very likely, a reminiscence of Augustine’s identification of the mental word
proper with an intellectual sight of some sort (visio cogitationis, De Trinitate
XV, 11, 20). 6

The second paragraph in Anselm’s quotation, on the other hand, further
characterizes the component units of mental speech properly speaking by mak¬
ing use of a cluster of features recognizably inspired by Aristotle’s description of
the “passions of the soul” in chapter 1 of the Peri Hermeneias (l6a5-8), which

5) See the passage quoted below n. 8. For a more detailed analysis of how Anselms ternary dis¬
tinction can be obtained from Augustine’s slightly more complex one, see Panaccio (1999),

156.

6) See also De Trinitate XV, 10, 18.

C. Panaccio / Vivarium 45 (2007) 269-282

275

Anselm was familiar with through Boethius’s translation and commentaries:
that these units are “the same for all”, namely, and that they are “likenesses”
(.similitudines ) of their external objects. Anselm, in these lines, equates Augus¬
tine’s mental words with Aristotle’s concepts or passions of the soul. Since
Augustine’s idea, historically, was to a very large extent independent from Aris¬
totelian souces, 7 what we have here, in the last analysis, is a clear case of a
tradition encounter, which in fact strikingly foreshadows Ockham’s own con¬
joining of Augustine and Boethius with respect to mental language in the first
chapter of the Summa logicae a couple of centuries later.

In Anselm’s case, however, the Aristotelian component is totally subservient
with respect to the Augustinian one. Just like Augustine’s, Anselm’s interest is
basically theological: he is striving to clarify what is going on in God himself.
He does use human psychology as a comparison for this, even recruiting Aris¬
totle in the process, but his goal is not to contribute anything to the study of
human thought; it is to tame to some extent the Christian idea of divine cre¬
ation. And the relevant chapter is organized accordingly, just as the entire
Monologion , indeed, follows a theological agenda. The key point of our specific
passage is made in the second sentence, when Anselm explains that the human
locutio mentis he wants to compare God’s creative thinking with, is not the
silent representation of linguistic signs, but the mental viewing of the things
themselves, the intent here being to pinpoint, for theological comparison, the
most spiritual aspect of human psychological activity, just as Augustine’s insis¬
tence on the mental word proper was basically aimed at improving our under¬
standing of the Christian God. Anselm, it must be stressed, is not merely
repeating Augustine: while Augustine was interested in the mystery of the
divine Trinity, Anselms primary focus in this passage is on the idea of the
creation ex nihilo , and he uses Augustine, therefore, in the context of his own
original undertaking, a situation which is quite typical as far as textual chains
are concerned. Yet, the meeting of the Aristotelian and the Augustinian tradi¬
tions in Anselm’s text is clearly unbalanced in the end: the Aristotelian refer¬
ence simply occurs as a secondary scientific support for what turns out to be a
generally Augustinian approach.

Albert the Great: “In I Sententiarum^ 27, 7”

In his Commentary on the Sentences , written toward the middle of the thir¬
teenth century, Albert the Great explicitly addresses the question of how to

7) See Panaccio (1999), especially chapters 2 and 3.

276

C. Panaccio / Vivarium 45 (2007) 269-282

reconcile the authorities as to the various senses of the term “verbum . This
occurs in distinction 27 of the first book, which is precisely the locus classicus
of thirteenth and fourteenth-century discussions on Gods Word and its com¬
parison with human mental speech. Albert, there, mentions four different dis¬
tinctions, all of which incorporate the theme of mental language:

- First, he quotes the very lines from Augustine’s De Trinitate , book IX,
which I have identified above as a possible source for Anselms division in
the Monologion; 8 and he extracts from it four varieties of verbum> “namely,
two with syllables, whether thought or pronuntiated, and two more cor¬
responding to the impressions of the external thing in the soul, according
to whether the thing is pleasant or not.” 9

- Second, he quotes a passage from John Damascenes De fide orthodoxy
book I, chapter 13, where four senses of “logos” — “verbum in the Latin
translation available to Albert—are enumerated: (1) the divine Verbum
“who is ever essentially present with the Father”; (2) “the natural move¬
ment of the mind, according to which it is moved and thinks and consid¬
ers, being as it were its own light and radiance”; (3) the “thought that is
spoken only within the heart”; and (4) “the utterance that is the messen¬
ger of thought”. 10

- Third, he mentions a commentary on John’s Gospel ( Super Ioannem)
where a threefold distinction is laid out between (1) the mental word
of the heart, (2) the internal image of the uttered word, and (3) the
spoken word.

- And finally, he attributes to some “masters” ( magistri) a similar distinc¬
tion between (1) the word of the thing itself, (2) the uttered word, and
(3) the mental representation of the uttered word ( species vocis).

The academic context, here, is very different from Anselm’s. The available tex¬
tual references had multiplied in the meanwhile and serious scholarship now
imposed new requirements. While Anselm could confidently assume that the
authorities he relied upon—and which he didn’t even bother to mention by
name—must have been in basic agreement with each other, just as reason and
faith were, this assumption has come to be a problem in Albert’s time. Albert,

8) Augustine, De Trinitate IX, 9, 15.

9) Albert the Great (1893), 46.

10) John Damascene (1955), 62-63. The English translation is from Salmond (1898), 16-17.

C. Panaccio / Vivarium 45 (2007) 269-282

111

it is true, insists that the four series of distinctions he is reviewing about the
verbum all amount to pretty much the same in the end (except that Augustine
draws a further distinction with respect to the others, between pleasant and
unpleasant cognitions). 11 But this supposed doctrinal equivalence is not some¬
thing he can afford to take for granted anymore; it now has to be argued for
with precise textual quotations and commentary.

Yet what is going on in Albert’s passage can be seen, just as in Anselm’s case,
as an encounter between the Augustinian tradition and the Greek philosophi¬
cal tradition. Augustine’s approach to the mental word is directly referred
to—and textually quoted—in Albert’s first division. It is, most probably, the
indirect source for the third and fourth divisions as well, which both display,
with a few terminological nuances, the same three-fold distinction we have
come across in Anselm between the mental word proper, the internal silent
representation of the spoken word, and the outwardly spoken word, a list
which yields an unmistakable Augustinian flavour. The second division, on
the other hand, is something completely different, with different origins.
Albert explicitly borrows it from Damascene, whose treatise, originally written
in Greek in the eighth century, was entirely independent from Augustine. The
exact sources of Damascene in the passage quoted by Albert are not known, 12
but the development is obviously rooted in a properly philosophical tradition.
Damascene’s work is a theological one, admittedly, but it often indirectly dis¬
plays philosophical sources. That this is the case in our present passage is
apparent from the following two considerations: first, the passage explicitly
makes use of the technical vocabulary of logos endiathetos (internal discourse)
and logos prophorikos (uttered discourse) which had become standard in
late Greek philosophy; 13 and second, once we put aside the first—overtly
theological—member of Damascene’s quadripartite distinction (the divine
Verbum “who is ever essentially present with the Father”), the remaining three
closely correspond to a ternary distinction attributed to the “Ancients” by the
Islamic thinker Alfarabi in the tenth century, who probably got it from philo¬
sophical sources. 14

10 Albert the Great (1893), 46-47.

,2) Cf. Panaccio (1999), 80-83.

13) For a detailed analysis of the distinction between logos endiathetos and logos prophorikos in late
Greek philosophy, see Panaccio (1999), 33-93. Contrary to a wide-spread belief, I argue there
(after a number of others, especially Chiesa 1991, 1992), that we have no good reason for think¬
ing that this distinction is of specifically Stoic origin. By the second century A. D., at any rate, it
had become common to all Greek philosophical schools.

14) See Alfarabi (1932), 136 (= De scientiis, ch. 2).

278

C. Panaccio / Vivarium 45 (2007) 269-282

As in Anselm, moreover, the two traditions are unequally present in Albert’s
passage. Not only is the Augustinian one granted three references against one
for the philosophical tradition, but the general context of discussion, most
importantly, remains typically Augustinian, the question under consideration
(“In reference to what earthly word is the divine Word so called?”) 15 being
clearly inspired by the De Trinitate. The unbalance, in this case, is further
interestingly reflected by the fact that Albert simply misinterprets Damascene s
distinction in the end by equating it with the other three. Damascene’s item
no 3 is—in so many words—the logos endiathetos of the Greek philosophers,
which cannot very well be interpreted as a mental representation of linguistic
sounds as Albert claims; and Damascene’s item no 2—the “natural movement
of the mind”, its “light and radiance”—, therefore, must be something quite
different from the internal discourse Albert wants to identify it with, 16 some¬
thing, in fact, that is utterly absent from Augustine’s division as well as from
the Super Ioannems and the magistris lists. Albert, in other words, misreads
Damascene because he reads him with Augustinian lens. To an even higher
degree than in Anselm, the encounter between the Greek philosophical tradi¬
tion and the Augustinian tradition leads in his text to an outright takeover of
the former into the latter. 17

William of Ockham: Summa logicae I, 1

Let us come back now to the text we started with from the first chapter of
Ockham’s Summa logicae :

It should be noted that just as Boethius says in his Commentary on the first book of the
Perihermeneias that there are three types of discourse [oratio] : written, spoken and concep¬
tual, the latter existing only within the intellect, there are, in the same way, three types
of terms: written, spoken and conceptual [...] And consequently, those conceptual terms
and the propositions which are composed of them are those mental words of which saint
Augustine says in book XV of the De Trinitate that they belong to no [external] language,
since they remain in the mind and cannot be uttered outside, although the spoken sounds
which are subordinated to them as signs are uttered exteriorly. 18

15) Albert the Great (1893), 46 (= In I Sent . 27, quest. 7).

16) Albert the Great (1893), 47.

17) The same Augustinian reinterpretation of Damascenes quadripartite division is also found in
Aquinas, who simply follows Albert on this. See Aquinas, In I Sent. 27, quest. 2, art. 1, and
Summa theobgiae I, 34, art. 1.

18) William of Ockham (1974), 7.

C. Panaccio / Vivarium 45 (2007) 269-282

279

Ockham, here, assimilates Augustine’s mental words with Boethius’s oratio
mentalis , thus bringing together, with respect to the theme of mental language,
the same two traditions—Augustinian and Aristotelian—that we have dealt
with in Anselm’s case. The operation, however, now takes on a completely
different significance.

From Boethius’s second Commentary on the Perihermeneias , Ockham
borrows the idea that there are three sorts of “ oratio ”. While not found in
Aristotle himself, the distinction was explicitly attributed to the later Peripa-
teticians by Boethius, and there is little doubt that it does belong to the
Aristotelian tradition. 19 Ockham uses it in what looks at first like a mere
comparison: “sicut secundum Boethium [...] triplex est oratio [...], sic triplex
est terminus' ’. The reference, however, is much more intelligibly seen as an
argument than as a comparison. “ Oratio ” is the technical term of the Latin
Aristotelian vocabulary for a complex expression, a complete sentence in par¬
ticular. Ockham’s point, consequently, is the following: since there are three
sorts of complex discursive units, as Boethius says, there must be, accord¬
ingly, the same three sorts of simple significant units. What comes out as
salient for Ockham in this context is the phenomenon of logical composition.
The conceptus terminus he arrives at on the basis of his Boethian premise is
first introduced by him as the basic component unit of propositional combi¬
nations: “The conceptual term is an intention or passion of the soul which
naturally signifies or cosignifies something, and which is destined to be a part
of a mental proposition ’. 20

When he turns to Augustine in the next sentence, Ockham continues to
express this same predominant interest for compositionality by explicitly
identifying Augustine’s mental words with “those conceptual terms and the
propositions which are composed of them". This is highly revealing. Augustine
himself never insisted very much on the compositional aspect of the mental
word. What he wanted to draw attention to was that the human mind engen¬
ders something spiritual within itself, and as a faithful expression of itself.
This is what matters, in his eyes, for the comparison he wants to push with
God’s generation of the Son. Ockham’s passage, by contrast, occurs at the
beginning of a treatise in logic , the general outlook and plan of which is
provided, precisely, by the compositional structure of both language and

19) See Boethius (1880), 29. A few pages further on (36), Boethius attributes the distinction to
Porphyry; cf. Panaccio (1999), 122-127.

20) Ockham (1974), 7.

280

C. Panaccio / Vivarium 45 (2007) 269-282

thought. 21 Its starting point is a specifically Aristotelian cluster of theses: logic
has to do with truth; internal thought is the primary locus of truth; and truth
is basically a matter of composition and division. It is Augustine’s role, now,
that has come to be ancillary in connection with the theme of the mental lan¬
guage: the main function of the reference to the De Trinitate in Ockham’s pas¬
sage is to legitimize, in the context of a generally Christian academic world, the
(non-Augustinian) idea that human thought has a semantical compositional
structure quite similar to what we observe in spoken and written languages.

One might be tempted to trivialize the shift: by remarking that Anselm’s
Monologion was a theological treatise while Ockham’s Summa logicae is a
philosophical work, and that it is normal, consequently, that the textual chain
rooted in Augustine’s De Trinitate should be predominant in the former while
the textual chain rooted in Aristotle’s Peri Hermeneias (via Boethius) is pre¬
dominant in the latter. But that would miss the point. What our examination
of the three chosen texts points to in the end is that the general significance of
the theme of mental language was deeply transformed from the late eleventh
century to the early fourteenth. It is not just a matter of one aspect of it being
more insisted upon in one field, and another aspect in another field. In so far
as the theme was present at all from Anselm to Albert, it was predominantly
dealt with in an Augustinian guise and in connection with prevailingly Augus-
tinian concerns. What happened between Albert and Ockham is that logic as
an autonomous field of study acquired a new salient institutional status in the
academic world and pressing questions about it came to the forefront. What
is logic all about? What are syllogisms composed of? And how are the basic
units of thought to be described. 22 The theme of mental language was radically
reoriented in this new context, mainly by Ockham, with the compositionality
of thought now at the centre of it, rather than the internal engendering of a
spiritual offspring. This shift from a theological to a philosophical exploitation
of the theme was associated, unsurprisingly, with a new balance in the use of
the relevant textual traditions, especially since the idea of the compositionality
of thought had itself clear Aristotelian rather than Augustinian roots. 23 The

21) The very first sentence of chapter 1 of the Summa logicae reveals the primary interest of the
work for compositionality, while, at the same time, announcing its general plan. See William of
Ockham (1974), 7: ‘Omnes logicae tractatores intendunt astruere quod argumenta ex proposi-
tionibus et propositiones ex terminis componentur.’ Part I, accordingly, deals with terms; Part II
with propositions; and Part III with arguments.

22) See on this development Panaccio (1992), (1999), ch. 8, and (2003).

23) See in particular Perihermeneias 1, 16a 10-13, and De Anima III, 6, 430a25-430b6.

C. Panaccio / Vivarium 45 (2007) 269-282

281

main thing, however, is that Ockham’s network of theoretical problems was
radically new with respect to both Anselm and Albert. His ultimate aim in the
Summa logicae is to use the technical apparatus of grammar and terminist logic
for the fine-grained analysis of the compositionality of human thought. This,
of course, is much more in line with the Aristotelian tradition than with the
Augustinian one. Yet it is, on the whole, a genuinely original research pro¬
gramme, even with respect to Aristotle and Boethius. And this research pro¬
gramme determines in the end the specific role played by each of the two
textual traditions in our chosen passage from the Summa logicae.

Conclusion

The three main channels through which the theme of mental language was
transmitted from the ancient to the medieval world were Augustine, Boethius
and John Damascene. How the corresponding three textual chains came across
each other in the Middle Ages with respect to mental language is revealingly
illustrated in the three passages we have looked at here. Anselm tacitly—and
casually—equates Augustine’s properly mental words with Aristotle’s concepts,
as known to him through Boethius. Albert the Great, with an impressive dis¬
play of scholarship, explicitly—but mistakenly—equates Damascene’s logos
endiathetos with Augustine’s silent representation of spoken words. And
Ockham, finally, equates Augustine’s mental word with Boethius’s oratio
mentalis. All three, then, concur in bringing together on this theme the
Greek philosophical tradition—the Aristotelian one especially—and the theo¬
logical Augustinian tradition. The outcomes, however, are heterogeneous.
While Anselm and Albert, despite their institutional distance, both generally
remain in the Augustinian line and exploit the philosophical reference as a
mere legitimizing device for theological purposes, quite the reverse occurs in
Ockham: Augustine serves there as a legitimizer for the promotion of a philo¬
sophical programme inspired by the Aristotelian tradition.

The whole thing, however, should not be seen as a large-scale continuing
struggle between two impersonal historical forces. The encounter in each case
is conducted with specific goals and interests, and within a distinctive context.
Anselm is primarily interested in understanding God’s creation ex nihilo.
Albert, while even closer to Augustine with respect to the problem he dis¬
cusses, stands out by the scholarly strategy he relies upon. And Ockham’s focus
on the compositionality of thought, while “inspired”, as I said, by Aristotelian
ideas, turns out in the end to be highly original in so far as it is intended at the

282

C. Panaccio / Vivarium 45 (2007) 269-282

exploitation of the specifically medieval terminist logic for the elaboration of
a new brand of philosophy of mind. Tradition encounters, in short, don’t just
happen, they are staged by individual authors—or groups of authors—who
use in their own ways and for their own purposes the intellectual and linguis¬
tic tools they find available in authoritative texts. Sometimes, of course, a
given author might subscribe on the whole to some ancient research pro¬
gramme, as Albert does with respect to Augustine’s problem of finding the
most adequate analogy between God’s internal engendering and human men¬
tal activity; but sometimes also, a new approach emerges, and it is usually
more illuminating for historians of thought in such cases to appreciate the
immediate relevant institutional context and to understand the philosophical
problems at hand than to locate the author’s remote textual sources.

BRILL

Vivarium 45 (2007) 283-297

VIVA

RIUM

www.brill.nl/viv

Intentionality and Truth-Making:
Augustine’s Influence on Burley and Wyclif s
Propositional Semantics’

Laurent Cesalli

Albert-Ludwigs-Universitat Freiburg i.Br.

Abstract

Walter Burley (1275-c. 1344) and John Wyclif (1328-1384) follow two clearly stated
doctrinal options: on the one hand, they are realists and, on the other, they defend a
correspondence theory of truth that involves specific correlates for true propositions,
in short: truth-makers. Both characteristics are interdependent: such a conception of
truth requires a certain kind of ontology. This study shows that a) in their explanation
of what it means for a proposition to be true, Burley and Wyclif both develop what we
could call a theory of intentionality in order to explain the relation that must obtain
between the human mind and the truth-makers, and b) that their explanations reach
back to Augustine, more precisely to his theory of ocular vision as exposed in the De
trinitate IX as well as to his conception of ideas found in the Quaestio de ideis.

Keywords

semantics, ontology, realism, intentionality, truth-making, medieval philosophy

Consider the two following questions: if propositional truth depends on a
kind of correspondence relation between the mental (i.e. a mental proposi¬
tion) and the non-mental (i.e. what that mental proposition is about), then
how do we bridge the gap between the mental and the non-mental? And
what kind of corresponding entity is responsible for the truth of so called non¬
standard propositions (i.e. propositions whose tense is not the present or
whose subject-matter belongs to the past or the future)?

0 I am very grateful to Prof. Maarten J.F.M. Hoenen and Prof. John Marenbon for their reading
of and critical comments on a first draft of this paper, as well as to Meredith Ziebart for her help
with the final English version of this study.

DOI: 10.1163/156833407X217777

284

L. Cesalli / Vivarium 45 (2007) 283-297

The present study intends to reconstruct the solutions that were given
to these problems by two prominent realist philosophers of the fourteenth
century, Walter Burley (1275-C.1344) and John Wyclif (1328-1384). These
solutions both display the remarkable common feature of all having been
elaborated under Augustinian influence. Two facts constitute the background
of this claim. First, when attempting to describe how what he calls a “real
proposition” {propositio in re) is formed, Walter Burley calls upon Augustine
and his theory of ocular vision as it is expounded in De trinitate XI, a theory
in which the notion of intentio plays a central role. 2 3 Second, when Burley and
Wyclif explain how one can generalize a rather strict correspondence-theory of
truth, they both draw upon Augustines Quaestio de ideis? Accordingly, my
aim is to show that reference to Augustine helped both Burley and Wyclif
with two issues crucial for propositional semantics, namely intentionality and
truth-making.

Following Aristotle but also Avicenna, most medieval philosophers, whether
realist or nominalist, defend a correspondence theory of truth, 4 whereby to be
true for a sentence is to correspond to something else. But not all medieval
philosophers understand correspondence in the same way. The question then
is: are there any proposition-specific correlates—we would say “states of
affairs”—that make propositions true? This is where nominalists and realists
disagree. The former deny the existence of such specific correlates, while the
latter, often speaking of an “ultimate propositional significate”, accept them. 5
Thus realist thinkers like Burley and Wyclif have to face the ontological chal¬
lenge of providing specific correlates, at least for every true proposition, 6 while

2) For an excellent study of the relation between optics and vision theories on the one hand, and
theory of knowledge and semantics on the other, see Tachau (1988).

3) For the medieval interpretation and reception of Augustine’s Quaestio de ideis, cf. Hoenen
(1997) and the 2004 issue of the “Revue thomiste”, entirely devoted to that theme.

4) Cf. for example Aristote, Metaphysics , IX. 10, 105lb 1-8 and Categories., 5, 4a34-37, 4b8-10
and 12, 14b 18-23. For the classical definition of truth as adaequatio intellects etrei and its origin
see for example Thomas Aquinas, De Veritate , q.l, a. 1 (1970), 5.162-6.200, as well as Muckle
(1933) for the origin of this definition by Isaac Israeli. For the avicennian definition, see Avi¬
cenna, Philosophiaprima, 1.8 (1977), 83.55.

5) Cf. for example John Duns Scotus, Quaestiones in primum et secundum librum Perihermeneias
Aristotelis (1639), q.2, n.4; Burley Quaestiones super librum Perihermeneias (1974), §3.553, Liber
Praedicamentorum (1497), f.l5vb-l6ra; Wyclif, Logice continuacio, 1.1 (1893), 77.

6) Falsity can be defined precisely as the absence of such correlate or with the correspondence to
an abstract (non real) « false » correlate, cf. Gregory of Rimini (1978-1987), III, 228.25-229.2.

L Cesalli / Vivarium 45 (2007) 283-297

285

nominalists can get around the difficulty by claiming that such correlates are
not needed at all. 7

The questions of intentionality and truth-making raised above are of central
importance for philosophical and theological issues: the philosophical issue is
that of the status and the possibility of scientific knowledge in a contingent
world; the theological one regards the relation between God and scientific
truth. Indeed, our problematic does not pertain to propositional semantics
alone, but to every domain of knowledge that is concerned with the status of
objects of propositional attitudes. If specific propositional correlates do exist,
they are not just ultimate significates or truth-makers of propositions, but also
objects of knowledge and belief—what p stands for in T know/believe that
p is always a proposition.

From the first decades of the thirteenth century onwards, medieval intel¬
lectuals had to face a conflicting situation: on the one hand, the Aristotelian
scientific paradigm as found in the Posterior Analytics requires all scientific
knowledge to be universal and necessary, 8 on the other hand, the created
world—that is: individuals, but also species and genera —is by definition con¬
tingent. Does this mean that God alone can be object and subject of genuine
scientific knowledge? To avoid this awkward consequence, one had to find a
solution that would be compatible with both the Aristotelian scientific exi¬
gency and the contingency of the created world.

Augustine’s theory of ideas as expounded in his Quaestio de ideis offered
such a solution. In this short text, Augustine turned Plato’s abstract Ideas into
ideas in the mind of God, thus providing, within the Christian framework, an
essential link between every contingent being and its eternal and necessary
counterpart. 9 The Platonic principle of participation made it possible for us to
have intellectual knowledge of the necessary and universal components of
contingent things. 10 This helped to provide a framework for “Aristotelian”

7) Cf. for example Adam Wodeham (1990), 195.19-30, who says that the question ‘ quid est
significatumproposition^?' an ill-formed question is ( quaestio inepta ), since what a proposition does
signify is not a quid , but an esse quid. Cf. also Gal (1967), 89 and Ockham (1974), 249-250.

8) Aristotle, Posterior Analytics, 1.4 (1968), 194.25-31). This text was translated around 1150 by
James of Venice. The first known latin commentary on the Posterior Analytics is due to Robert
Grosseteste and is dated around 1220-1230, cf. Robert Grosseteste (1981), 18-9.

9) Augustine, Quaestio de ideis (1975b), 70.1-71.32. For the medieval reception and tradition of
Augustine’s Quaestio de ideis, cf. Hoenen (1993), 121-156 and (1997), for literature, see pp. 245-246,
n. 3-4, as well as Hoffmann (2002).

10) Augustine, Quaestio de ideis (1975b), 71.33-73.64.

286

L. Cesalli / Vivarium 45 (2007) 283-297

science in a contingent—that is created—world, a framework that was recog¬
nized and used by Burley and Wyclif in their propositional semantics. However,
there arose a further, theological problem: if, as realist philosophers main¬
tained, (i) true scientific propositions are made true by corresponding entities
and (ii) all genuine scientific knowledge has to be universal and necessary,
then the truth-makers of scientific propositions are universal and necessary
entities subsisting alongside God, since not every scientific truth pertains to
God. This was theologically problematic insofar as the view that there are
eternal truths distinct from God had been repeatedly discussed and con¬
demned by the church. 11 Thus, as we will see, realist philosophers like Burley
and Wyclif are careful to develop their ontology without threatening Gods
uniqueness and simplicity.

Intentionality

Wyclif and the Truth That Pertains to the Logician

Discussing Book IV, chapter 6 of Aristotle’s Metaphysics , Wyclif gives a four¬
fold division of truth: truth, he says in his De ente (written between 1365-
1375), is sometimes understood in a semantic way, that is as the correspondence
by virtue of which a linguistic sign is true or false; sometimes, truth is under¬
stood in a semiotic way, that is when the sign itself is called a truth; 12 some¬
times, it is understood is an ontological way, when extramental reality is called
a truth; finally, truth is understood in an intermediate or mixed way, that is as
an aggregate made out of ontological truth and a mental act: “ aggregatum ex
veritate reali et actu anime” u Now Wyclif adds,

n) The condemned thesis reads “Quod multae veritates fuerunt ab aeterno, quae non sunt ipse
Deus”. The series of condemnations starts in 1241 with William of Auvergne, we find it men¬
tioned in Bonaventures Sentence Commentary (II, d.23, a.2, q.3) and it figures in the Parisian
and Oxonian lists of errors condemned in 1277 respectively by Etienne Tempier and Robert
Kilwardby. For a reconstruction of this intricate story, cf. De Libera (2002), 177-187.

12) The distinction between the semantic and the semiotic truth may be understood analogically
to the distinction made by Anselm of Canterbury in De veritate , 2 between two kinds of propo¬
sitional truth: “alia igitur est rectitudo et veritas enuntiationis, quia significat ad quod significan-
dum facta est; alia vero, quia significat quod accepit significare” (Anselm (1946), 179.10-12). For
a more detailed account of Wyclif s theory of propositional truth, see Cesalli (2003).

,3) Wyclif (1909), 103.

L. Cesalli / Vivarium 45 (2007) 283-297

287

[. • •] in order to explain the fourth kind of truth, one has to pay attention to the common
opinion of the doctors, which is that some things are entirely outside the soul, such as
wood, a stone, and the like; others are entirely within the soul, such as dreams and acts of
the soul; however some things, they say, are partially inside and partially outside the soul,
such as a universal, time, as well as other aggregates resulting of the operation of the intel¬
lect and extra-mental things. This is why philosophers say that the intellect produces uni¬
versality in things. 14

We shall see below that Walter Burley is most probably one of the Doctores
Wyclif has in mind when he refers to a vulgata sentencia doctorum. According
to Wyclif, the four kinds of truth can be distributed to different disciplines.
Thus, the grammarian is interested in semiotic truth and the metaphysician in
ontological truth. As for the first and the fourth kind of truth, they pertain to
the logician 15 who considers the truth in a mixed way {mixtim) y that is as an
ontological truth as it is considered by the mind: “Thus the logician conceives
in a mixed way that the proposition ‘the world exists’ is true on the side of
things in so far as the mind thinks about it”. 16 Summing up, truth as it per¬
tains to the logician results from a mixed causality of extramental or ontologi¬
cal truth on the one hand, and a cognitive mental act, on the other. In modern
terms, one could say that a true proposition is, according to the logician, an
extra-mental truth present in the mind as a complex intentional object.

Burley and the “ens copulatum”

Burley introduces the notion of propositio in re in his so called middle com¬
mentary on the Perihermeneias (c.1310) to answer the following relativist
objection: if the truth of a proposition is primarily in the mind, does this
mean that there will be as many truths as there are minds composing them?
Not at all, replies Burley. What is multiplied according to the number of

14) Ibid.: “Unde pro explanacione quarti membri, notanda vulgata sentiencia doctorum, que est:
aliquod est totaliter extra animam, ut lignum, lapis, et cetera; aliquod totaliter in anima, ut
sompnia et actus anime; aliquod autem, ut dicunt, partim in anima et partim extra animam, ut
universale, tempus et cetera aggregata ex operacione intellectus et rebus extra. Unde dicunt phi-
losophi quod intellectus facit universalitatem in rebus”. For the authority according to which the
intellect causes universality in things, cf. Auctoritates Aritsotelis, 6.27 (Hamesse (1974), 176), the
source of which is Averroes, In De anima I, com. 8 (Averroes (1953), 12.25-26).

15) Wyclif (1909), 108: “Illud <sc. ‘verum et falsum non sunt in rebus, sed in mente’> verum est
de vero et falso primo et quarto modo dictis, quorum consideracio pertinet ad logicos, qui non
consistunt citra composicionem et divisionem in anima”.

16) Wyclif (1909), 109: “logicus autem concipit sic mixtim, quod ‘mundum esse est verum ex
parte rei, ut mens cogitat”.

288

L Cesalli / Vivarium 45 (2007) 283-297

individual intellects are mental subjective propositions, but their truth is itself
dependent on their correspondence with an objective mental proposition that
one can call, so says Burley, a propositio in re. Such an objective mental propo¬
sition is the ultimate propositional significate. Many years later, in his last
commentary on the Categories (1337), Burley expounds his theory once again,
and this time he is more careful to explain exactly what that ultimate proposi¬
tional significate is. Let us now consider two features of this propositio in re.

First, it is described as being partly ( partim ) in the mind and partly outside
the mind; more precisely, its formal part (the copula) is a mental act, while its
material parts (subject and predicate) are extramental things. 17 A similar thesis
is already present in Burley’s early Quaestiones super Porphyrium , which can be
read as a synthesizing lecture on the last two cases of the Avicenno-Albertinian
distinction between universal ante rem, in re , and post rem. ]8

Second, when it comes to the question of how such a hybrid item can pos¬
sess any kind of unity, Burley calls upon the De trinitate , explaining the forma¬
tion of the propositio in re according to the model of vision analysed in book
XI of Augustine’s theological treatise: 19 just as the sense of vision possesses a
force called intentio , which is capable of maintaining the sense of vision cou¬
pled with a visible object, thus producing the actual vision of an extramental
object, so the intellect possesses an intentio capable of producing the actual
signification of an object:

But one might ask here how it is that a single composite can be made from a thing existing
in the intellect and a thing existing outside of the soul. To this it must be answered that

17) Burley (1497a), I6rb: “Unde cum propositio sit triplex, quaedam in prolatione, quaedam in
conceptu et quaedam significata per propositionem in conceptu que potest dici propositio in re,
propositio primo modo dicta, scilicet propositio in prolatione, est totaliter extra animam et talis
propositio totaliter componitur ex vocibus que habent esse extra animam. Propositio vero com-
posita ex conceptibus est totaliter in intellects Et compositio composita ex rebus partim est in
intellectu et partim extra intellectum. Quantum ad suum formale est in intellectu sed quantum
ad materialia est totaliter extra intellectum”.

18) Burley, Quaestiones super Porphyrium , q. 7, quoted in Burley (1998), 78, n.18; Albert the
Great, Depraedicabilibus , II.3, (1889), 24a-b; Avicenna (1308), f.l2ra. Avicenna himself takes
this doctrine from the neoplatonic Commentators of Aristotle such as Ammonius and Simpli¬
cius. It ultimately goes back to Plato and Aristotle: the universal ante rem corresponds to the
Platonic Idea, the universal in re to the Platonic participated form and the Aristotelian imma¬
nent form, the universal post rem to the abstract concept according to Aristotle, see De Libera
(1996), 183-183, 501.

19) Augustine, De trinitate , XI, ii, 2. For the role of the notion of intentio in Augustine’s theory
of ocular vision and mental language, see Sirridge (1999).

L. Cesalli / Vivarium 45 (2007) 283-297

289

such a composite can be made by intellectual but not by real composition, and that such a
composite can be called a copulated being; and a copulated being can be produced not only
by the intellect, but also by the senses or by the intention of the senses. For Augustine says
that intention couples the sense, or the operation of the sense, with the sensible object. 20

In its most articulate and final formulation, then, Burleys theory of the propo-
sitio in re is explained along the lines of Augustine’s analyses of ocular vision.

Both Burley and Wyclif describe a kind of logical object grounded in extra¬
mental reality as well as in mental activity. The former explains its formation
in reference to the theory of vision as found in the De trinitate , the latter
claims he is just reproducing a vulgata sententia doctorum , one of these doctors
most probably being Burley himself. 21 A propositio in re , says Burley, is partly
in the soul and partly outside of the soul, while according to Wyclif, the truth
as it pertains to the logician is conceived in a “mixed way”, that is: when a
cognitive act is directed towards an extramental truth. In both cases, what is
described is the ability of the mind to produce complex intentional objects.

Truth-Making

Burley and the Permanence of Truth

Burley defends a correspondence-theory of truth: a proposition is true if some¬
thing corresponds to it and makes it true. More precisely, what makes a prop¬
osition true is the identity of the things signified respectively by the subject
and predicate terms. Such a theory works well in simple, factual cases. But
how about when the things which make a proposition true no longer exist?
Burley considers and provides a solution to this kind of problem: the proposi¬
tion ‘ Caesar est Caesar , for example, is true now because there is now a relation
of identity between Caesar and himself, not as a “being as it exists”, but as a

20) Burley (1497a), I6rb-va: “Sed dubium est hie qualiter potest fieri unum compositum ex re
existente in intellectu et re existente extra animam. Dicendum quod ex talibus potest fieri unum
compositum compositione intellectuali non autem compositione reali et tale compositum potest
dici ens copulatum et potest fieri ens copulatum et non solum per intellectual sed edam per
sensum vel per intentionem sensus. Dicit enim beatus Augustinus quod intentio copulat sensum
vel operationem sensus cum sensibili obiecto”.

21) Besides the patent similarities in Burleys and Wyclif s respective theories of the ens copulatum
and of the veritas quae pertinet ad logicum, the claim that Wyclif knew Burleys theory
of the proposition is supported by its explicit mention in the De universalibus , 1 (Wyclif
(1985a), 21.85-92).

290

L. Cesalli / Vivarium 45 (2007) 285-297

“super-transcendent being” or a “being that is common to all intelligibles”. 22
Does this mean that anything one can think of can be the truth-maker of a
proposition? Burley’s insistence on the objective character of truth is certainly
incompatible with such an interpretation. So what is that “being common to
all intelligibles”? Let us try to clarify this point by considering some of Burley’s
other texts.

The problem here is what one can call the problem of the permanence of
truth, that is, of the permanence of the two extremes of the correspondence
relation which defines truth: a proposition and its truth-maker. To be able
to deal with non-simple factual cases, Burley introduces into his theory
an independence of being and actual existence on three different levels. We
have just encountered the first with the example of the proposition ‘Caesar est
Caesar —the independence of the being of a truth-maker in regard to its actual
existence via the notion of “being common to all intelligibles”. The second is
the independence of ontological truth in regard to the intellect. The third is
the independence of the truth of any proposition in regard to its actual forma¬
tion. All of these elements occur in the following text, taken from the so called
middle commentary on the Perihermeneias , probably written around 1310:

Thus I say that the thing signified by the proposition man is an animal’ does not depend
on the intellect, just as the truth of that thing does not depend on it; (...) The mental
propositions that the intellect makes by perceiving such extra-mental truths correspond to
things that are so related in reality. Thus I say that the truth which is subjectively in the
intellect is nothing but a certain correspondence between the intellect and a proposition
that is only objectively in the intellect. (...) Therefore, it is possible for the proposition ‘man
is an animal’, insofar as it is subjectively in the intellect, not to exist, but insofar as it is
objectively in the intellect, it is impossible for it not to be, or not to be true. 23

22) Burley (1497b), 58rb: “Dicendum quod Caesare corrupto identitas est Caesaris ad Caesarem,
sed ilia identitas non existit, sed est identitas rationis. Et idem Caesarem idem Caesari identitate
quae non est nec oportet quod idem et diversum semper sint differentie entis maxime transcen-
dentis quod scilicet est in intellects Unde sic potest dici, quod ens dicitur dupliciter: uno modo
ut est commune omni intelligibili, alio modo idem est quod existens. Sic idem et diversum dici¬
tur uno modo ut sunt differentie entis transcendentis, alio modo ut sunt differentie entis in
effectu, hoc est in actu existentis”.

23) Burley (1973), §1.27: “Unde dico quod res significata per istam ‘homo est animal’ non
dependet ab intellectu nec etiam veritas isdus rei; (...) Istis tamen sic se habentibus in re corres¬
pondent propositiones in intellectu quas intellectus efficit ex hoc quod percipit tales veritates
extra. Unde dico quod veritas quae est subiective in intellectu non est nisi quaedam adaequatio
intellectus ad propositionem veram quae solum habet esse obiectivum in intellectu. (...) Unde
ista ‘homo est animal’ quae est propositio habens esse subiective in intellectu potest non esse,
tamen ista ut solum habet esse obiectivum in intellectu non potest non esse nec non esse vera”.

L Cesalli / Vivarium 45 (2007) 283-297

291

We find the very same idea in Burleys Super tractatum fallaciarum (probably
written at Oxford, in the first decade of the fourteenth century). 24 There, Bur¬
ley considers an objection that is similar to the one mentioned above and in
reply to which he had introduced the notion of propositio in re. In the Super
tractatum fallaciarum , the problem is the following: Burley has just stated that
insofar as a syllogism exists in the mind {in mente ), it cannot cease to be, as
opposed to what happens to written or spoken syllogisms. Now comes the
objector: if the proper being of a syllogism is in the mind, then, when a syl¬
logism is no longer in the same mind—for example not in my mind anymore,
but in someone else’s—it must cease to be; and thus, Burleys thesis of the
permanent existence of a syllogism in the mind is mistaken. Burleys answer is
surprising:

Against this it is argued that if the proper being of a syllogism were in the mind, then it
would cease to be if at some later point it were no longer in the same mind (...). To that, it
must be replied that the being of a proposition or syllogism does not depend essentially
upon my mind or yours, but upon the intellect of an intelligence from which it never
recedes. Thus if a syllogism were now in a particular mind and then later not, it would cease
to have the subjective being that it had, but would not cease to be essentially. 25

There are two points of particular interest in this passage: the first is the expla¬
nation of the permanence of a proposition through a non-human intellect,
that is, through the intellect of “some intelligence” {intellectus alicuius intelli-
gentiae)? G the second is the link implicitly made here between the objective
being of a proposition and the intellect of that intelligence. In other words,
when Burley says of a proposition as being objectively in the soul that it is
impossible for it not to be, nor to be untrue (cf. Burley (1973), §1.27 cited

24) Burley (2003), 131-207 (197-207).

25) Burley (2003), 200: “Contra illud arguitur: si proprium esse syllogismi esset in mente, si
postea non sit in eadem mente, desinit inesse (...). Ad illud dicendum quod esse propositionis
vel syllogismi non dependet essentialiter a mente mea vel tua, sed essentialiter dependet ab intel-
lectu alicuius intelligentiae a quo numquam recedit. Unde si aliquis syllogismus nunc sit in
mente alicuius et postea non, desinit habere esse subiectivum quod habuit, non tamen desinit
esse essentialiter”.

26) As S. Ebbesen remarks (Burley (2003), 136): “It is no common place in works of logic to
appeal to the intellect of an intelligence, i.e. of a separate substance, in order to provide proposi¬
tions with a permanent essence”. For the incorporation of the (originally Aristotelian) intelli¬
gences, that is intellectual and non-material substances, into medieval latin Philosophy, see for
example Thomas Aquinas, De ente et essentia , c. IV. Besides Avicennas Metaphysica or Philosophia
prima , the neoplatonic Liber de Causis plays a prominent role in that process of integration.

292

L. Cesalli / Vivarium 45 (2007) 283-297

above p. 290), he might mean its being objectively contained in the intellect
of a non-human intelligence. 27

In this respect, it is worth comparing what Burley says in the above cited
texts with what can be found in his later metaphysical treatise De universalibus
(written after 1337). The whole fifth chapter of that treatise can be seen as a
commentary on Augustine’s Quaestio de ideis. There, Burley asks the following
question: “whether <universals> exist in God, as those say who assert that the
divine mind contains ideas representing the images of created things, or
whether they exist per se outside of the divine mind”. His answer is that uni¬
versal, or ideas, do exist in the divine intellect, but one has to be careful not
to identify simpliciter ideas with the divine essence. 28 An idea, according to
Burley, is a kind of connotative entity. In its “principal” dimension, an idea in
the divine mind is really identical with the divine essence; in its connotative
dimension however, it is different from it. For example, in their “principal”
dimension, that is, as pre-representations of creatures, the ideas both of man
and horse are identical with the divine essence; but the connotata (Burley also
speaks of ydeata) of those ideas, that is, the creatures man and horse, are dis¬
tinct from the divine essence:

Therefore Augustine says in his book On 83 Questions , that they <i.e. ideas> were from
eternity exemplary models representing things to be created (...). (...) the idea of man and
the idea of horse are really identical, because they are the same as the divine essence; never¬
theless, they are different with respect to their connotation, because the idea of man con¬
notes man and not horse, and the idea of horse connotes horse and not man. 29

27) In question 4 of his Quaestiones super Sophisticos Elenchos (Burley (2003), 162), Burley also
mentions a non-human intelligence in his argumentation. As a reply to an objection according
to which a vocal sound {vox) cannot be equivocal because, as a bearer of forms, a vocal sound
cannot bear at the same time different forms of the same kind (in the present case, different
rationes significandi) , Burley states that this would be true if the rationes significandi were real
forms like whitenesses for example. However, the rationes significandi at stake are not real but
intentional forms (formae intentionales) which can exist simultaneously in the same subject as,
for example, a vocal sound. To reinforce his answer, Burley describes the case of the separate
intelligences “who contain eternally all the species of intelligible things”.

28) Burley (1998), 32: “Ex istis potest patere error quorundam dicencium quod omnes res intel-
lectae a deo ab eterno sunt idem quod essencia divina”.

29) Burley (1998), 46-48: “Et ideo ponit Augustinus in libello suo De Octoginta Tribus Ques-
tionibus quod ab eterno fuerunt <sc. Ydee> exemplaria representancia res faciendas. (...) ydea
hominis et ydea equi sunt idem realiter, quia sunt idem quod essencia divina, sunt tamen diversa
quantum ad connotata, quia ydea hominis connotat hominem et non equum, et ydea equi con-
notat equum et non hominem”.

L. Cesalli / Vivarium 45 (2007) 283-297

293

Now, as things to be created the connotata ox ycleat a horse and man are eter¬
nally and objectively present in the divine mind, and are distinct from the
divine essence only in an intelligible way. 30

We can now, I think, try to bring together the different pieces of the puzzle:
the problem of the permanence of truth, that is, of the generalisation of a
correspondence-theory of truth, is solved by Burley thanks to the notion of
the objective being of universal natures, propositions and syllogisms, in the
precise sense of their being permanent objects of permanent intellection by a
non-human intellect. In his earlier logical works, Burley speaks of intelligentia ;
in his later metaphysical tract on universal, he speaks of the divine mind and
explicitly links his solution of the problem with his interpretation of Augus¬
tine’s Quaestio de ideis . Coming back to the example we started from, we can
now say that what grants the truth of a proposition like ‘ Caesar est Caesar
where Cesar is not alive anymore, is the identity of Cesar with himself as
objectively present in, that is, eternally thought by, a non-human intellect. 31

Wyclif and Divine Truth-Making

There is a very strong truth-making principle in Wyclif’s thought. As Antony
Kenny puts in the notes of his translation of the De universalibus (c.1374):
“According to <Wyclif>, predication <is> not just the attachment of a predi¬
cate to a subject within a sentence, but the feature of the real world which
makes such a predication true.” 32

As we did in Burley, we also find in Wyclif the idea that truth possesses a
kind of permanence by being permanently the object of intellection of some
intellect—human, or non-human. In his De ente —a metaphysical work already
mentioned when examining Wyclif’s typology of truth 33 —he states that if one
wants to know which of the four kinds of truth depend on an intellect, and

30) Burley (1998), 50-54: “Dicendum quod ydeata fuerunt eterna obiective in essencia divina,
hoc est ab eterno erant representata et intellecta a deo. (...). Concedo quod res creanda est aliud
a deo accipiendo ly ‘aliud’ ut est differentia ends transcendentis; et eciam quoad ens accipiendo
ly ‘ens’ ut est nomen maxime transcendens, quia ens isto modo dictum est convertibile com hoc
communi ‘intelligibile’”.

3,) Similar theories of the permanence of truth are found in Augustine’s Soliloquia (II.2), Greg¬
ory of Rimini’s Lectura (1978-1987), t. 3, 228 and in the Abreviatio of Adam Wodeham’s Lectura
by Henry Totting of Oyta (1512), 118r.

32) Wyclif (1985b), 182.

33) See above, p. 286.

294

L. Cesalli / Vivarium 45 (2007) 283-297

which do not, one has to say that each kind of truth depends on an intellect,
provided we do not restrict the notion of intellect to the human one. 34

Thus, the semantic and semiotic truths of a sentence, the truth of a judg¬
ment as considered by a logician and the ontological truth of reality all are
dependent on an intellect: the first three depend immediately on the human
intellect, the fourth on the divine intellect. But all four depend ultimately on
the divine intellect, and this for two reasons: 35 first, because whatever is cre¬
ated or caused exists primarily in the divine intellect; second, because the
divine intellect is responsible for the intelligibility of the world with respect to
our intellectual faculty as well as its possible objects. As for necessary truths
like mathematical propositions, they depend only on the divine intellect.
Indeed, says Wyclif, if all intellects were destroyed—read: all human intel¬
lects—God would still know that 2 plus 3 equals 3. 36 Still, and even if there
cannot be any human science capable of giving an accurate account of such
truths per se, they nevertheless have a cause:

And thus there cannot be a science that would be adequate to those sacred things that are
analytical truths. Therefore, Augustine writes in De doctrina Christiana book 2, chapter 32:
‘that truth of connections has not been instituted, but only noticed by human beings, since
it is in the eternal nature of things by divine institution. 37

We, as human, can grasp these divine truths, not because we are able to see
what is in the divine mind, but because the intelligibility of created things
somehow provides us with access to those eternal truths. Now if God, or the
divine mind, is the ultimate cause of such necessary truths, one must be careful
on one hand, not to identify them simpliciter with God (because they are many,
and God is one), and on the other hand, not to separate them simpliciter from
God (because if anything is eternal, then that thing can only be God):

34) Wyclif (1909), 103-106.

35) Wyclif (1909), 108.

36) Wyclif (1909), 103. Although mathematical true propositions like ‘2 + 3 = 5’ are not onto¬
logical truths in the same sense as, for example, the existence of Socrates is one (if Socrates exists),
they share a common property with the ontological truth, namely their subsisting independently
of any human intellect. Furthermore, Wyclif suggests a connection between mathematical and
ontological truth.

37) Wyclif (1909), 105-106: “Et tamen inpossibile sit dari sciencia proporcionata illis sanctis que
de virtute sermonis sunt veritates. Unde Augustinus, 2° De doctrina Christiana 32°, sic scribit:
‘ilia veritas coneccionum non instituta est, set animadversa ab hominibus, cum sit in rerum
racione perpetua et divinitus instituta”. For the reference to Augustine, see De doctrina Christi¬
ana , 11.32.

L Cesalli / Vivarium 45 (2007) 283-297

295

(...) no such truth, although it has existed forever, expresses an essence or a nature which is
eternal in such a way that it can only be God. (...) Nevertheless, formally speaking, there is
no doubt that they are mutually distinct, as well as distinct from God, as we read in Augus¬
tine’s question 46 De ydeis.™

Necessary or eternal truths are in God just as other res rationis are in Him, that
is: without having to be really distinct from Him. There is a real identity and
a formal distinction in God between its essence and, for example, eternal
truths. 39 Making God’s intellect the place where necessary or eternal truths
subsist, is not very original. But Wyclif goes one significant step further: in
addition, contingent truths like possibilities, past and future truths also subsist
in the divine intellect. Moreover these truths are said to be signified by the
corresponding propositions. 40

Certainly, when we utter a true proposition such as, for example, ‘ Sor est
homo , we are not talking about God’s intellect nor about its content. But if
it so happens that the proposition ‘ Sor est homo is true, what makes it true,
ultimately is in the divine intellect. As Wyclif puts it in the De universalibus ,
commenting Augustine’s Quaestio de ideis:

And if it is asked: what is such an ability to produce <truth>? It is replied that it is nothing
which differs essentially from God. Thus, formally, it is not something but a thing of rea¬
son, it does not belong to any created genus but it is present by reduction in the same genus
in which is its ideatum, since it is its principle, just as the idea of human being is in the
human species, and consequently in the genus of substance, and in a way, speaking about
the species, it is per se the species of human being. 41

What makes propositions true, regardless of their contingency or necessity, are
the intelligible traces left by God’s Ideas in creation. 42 The Ideas themselves are

38) Wyclif (1909), 106: “(...) nulla talis veritas, quamvis sit eterna a parte ante, dicit aliquam
essenciam vel naturam sic eternam, nisi solum deum. (...) Et tamen, loquendo formaliter, non
<est> dubium quin distinguntur ab invicem et a deo, ut patet per Augustinum Deydeis questione 46”.

39) Wyclif (1909), 112.

40) Wyclif(1909), 111.

41) Wyclif (1985a), 373: “Et si quaeratur: quid est talis productibilitas? Dicitur quod nihil essen-
tialiter nisi Deus. Formaliter autem non est aliquid sed res rationis, non per se in aliquo genere
creato sed per reductionem in eodem genere in quo est suum ideatum cum sit eius principium,
ut idea hominis est in specie humana, et per consequens in genere substantiae, et uno modo,
loquendo de specie, est per se species hominis (...)”.

42) In the following lines of the text quoted in the previous note, Wyclif calls upon Robert Gros¬
seteste and his own conception of the relation between divine ideas and universal as exposed in his
commentary on Aristodes Posterior Analytics, 1.7, Cf. Robert Grosseteste (1981), 139.96-142.157.

296

L Cesalli / Vivarium 45 (2007) 283-297

in the divine mind, but what human beings can grasp directly and start from
in their intellectual activity, are in the created things themselves as ideata.

Conclusion

Let us conclude with three points. First, this short study of the Burleyan prop-
ositio in re and the Wyclifian veritas quae pertinet ad logicum has shown that
Wyclif’s approach is dependent upon Burleys, whose model is in turn Augus¬
tine’s description of ocular vision in the De trinitate. The common preoccupa¬
tion of these three thinkers is with an explanation of how the mind, interacting
with the extra-mental world by means of its intentio , produces specific, inten¬
tional objects: the content of sense perception, for Augustine, the truth which
pertains to the logician, for Wyclif, and for Burley, the ultimate significate of
propositions. 43

Second, when, in the context of correspondence-theories of truth as we
have them in Burley and Wyclif, it comes down to the question of how
such theorie work in cases in which no actual signifcates exist—as in the case
of true propositions about non-existent objects, or in the case of eternal
truths—both Burley and Wyclif call upon a non-human intellect as the place
where that which can be expressed by human sentences subsists as a perma¬
nent object of thought; and they do so in reference to Augustine’s Quaestio
de ideis.

Third, as for the specific question of the permanence of a proposition which
is not actually formed by anyone, Burley seems to rely, on at least two occa¬
sions, on an Avicennian-like cosmological gnoseology. In spite of the fact that
the terminology and the dialectical context of these passages are similar to
those in which Burley uses Augustine’s Quaestio de ideis , one cannot, I think,
simply replace ‘ intelligentid in the earlier texts by ‘mens divind —one sufficient
reason for that being that the plural ‘ intelligentiis used there by Burley is
incompatible with divine simplicity.

This brief analysis of Burley’s and Wyclif’s solutions to the crucial problems
of ( a ) giving an explanation of mind/world interaction and ( b ) generalizing
a strict correspondence theory of truth, has shown that Augustine played an
important role in such matters in the fourteenth century. Realist thinkers like

43) For the role played by Augustine not only in medieval theories of intentionality, but also
in Franz Brentanos characterization of intentional objects as possessing a so called mental in¬
existence, see Perler (2002), 403-405.

L. Cesalli / Vivarium 45 (2007) 283-297

297

Burley and Wyclif found in Augustine’s works, besides his definition and
classification of signs, and theory of the inner word, constitutive elements for
the semantic and ontological aspects of their propositional logic. These ele¬
ments became constituent of what we may call their theories of intentionality
and truth-making.

BRILL

Vivarium 45 (2007) 298-310

VIVA

RIUM

www.brill.nl/viv

Names That Can Be Said of Everything:
Porphyrian Tradition and ‘Transcendental’
Terms in Twelfth-Century Logic*

Luisa Valente

La Sapienza Universita di Roma

Abstract

In an article published in 2003, Klaus Jacobi—using texts partially edited in De Rijk’s
Logica Modernorum —demonstrated that twelfth-century logic contains a tradition of
reflecting about some of the transcendental names ( nomina transcendentia) . In addi¬
tion to reinforcing Jacobi s thesis with other texts, this contribution aims to demon¬
strate two points: 1) That twelfth-century logical reflection about transcendental terms
has its origin in the logica vetus , and especially in a passage from Porphyry Isagoge and
in Boethius’s commentary on it. In spite of the loss of the major part of the Aristotelian
corpus, the twelfth-century masters in logic still received some Aristotelian theses con¬
cerning the notions of one and being via Porphyry and Boethius; on the basis of such
theses, they were able to elaborate a sort of proto-theory of the transcendental as
trans-categorical terms. 2) That this theory is centred on the idea that there exists a
particular group of names which have the property that they can be said of everything;
this group includes “being”, “one”, “thing” and “something” ( ens , unum, res , aliquid).
Twelfth-century masters in logic try to question the (originally Aristotelian) thesis that
these terms are equivocal, although they do not deny it completely.

Keywords

transcendental terms, one, being, thing, something, Porphyry, Boethius, Peter Abelard,
twelfth-century logic

In his book on the medieval theory of transcendental published 1996, Jan A.
Aertsen—following D. H. Pouillon—suggests that the first mature exposition

I am grateful to Frosty Loechel for assisting in the English translation. Particular thanks to
Riccardo Chiaradonna, Sten Ebbsesen, and Alfonso Maierii for the very useful information and
suggestions they provided. Any mistakes or imprecisions are of course my own responsibility.

DOI: 10.1163/156853407X217786

L. Valente / Vivarium 45 (2007) 298-310

299

of this theory is the presentation of the prima and communissima ens, unum,
verum, bonum contained in the Summa de Bono written by Philip the Chan¬
cellor around 1225-1228. 1 As sources of this theory Aertsen suggests princi¬
pally the new Aristotle ( Metaphysics and Nicomachean Ethics) as well as writers
in Arabic (mainly Avicenna). 2 The studies which followed the book confirm
this interpretation, though some of them have called attention to the role
played by thirteenth-century logic in the constitution of the theory of tran-
scendentals. 3 Some researchers asserted the existence, in the Latin world before
the impact of the Greek-Arabic philosophy, of elements which go in the direc¬
tion of the later theory of transcendentals; but they stress that such elements
were isolated and occasional. 4

In an article published in 2003, however, Klaus Jacobi demonstrated that
twelfth-century logic contains a tradition of reflecting about some of the tran¬
scendental names ( nomina transcendentia) . 5 In his article Jacobi used texts pre¬
sented and partially edited in L. M. De Rijk Logica Modernorum\ consequently
in this paper I shall consider almost exclusively twelfth-century logical texts
not included in De Rijk’s pioneering work.

In addition to reinforcing Jacobi’s thesis with other texts, my contribution
aims to demonstrate two points:

1) First of all, that twelfth-century logical reflection about (some of the)
transcendental terms has its origin in the logica vetus , and especially in a
passage from Porphyry Isagoge and in Boethius’s commentary on it. The

Aertsen (1996); Pouillon (1939). For a general review of the literature concerning medieval
theory of transcendentals, see Aertsen (1991, 1999). See also Marenbon (1992), Ventimiglia
(1998), Aertsen (2000), Pickave (2003)—with a bibliography of Aertsen’s studies—, Valente
(2003) and (2006).

2) Aertsen (1996), 4l9ff. See also Gracia (1992), 117f., who mentions Aristotle’s Metaphysica
993b30, 1003b25, 1004b 15, 1051a34, 1061 a 15, and Ethica Nicomachea 1096a23, and, about
the Arabic sources, De Libera (1994).

3) Cf. Ventimiglia (1998), 526, Aertsen (2000), and Pini (2003). De Libera (1994) had already
noticed the role of 13th century logic.

4) See for example Gracia (1992), in part. 117ff., Marenbon (1992), S. MacDonald (1991) and
(1992).

5) “Transcendent names” (“nomina transcendentia”)—a term which is usd in the treatise Dialec-
tica Monacensis\ see Jacobi (2003), 34—seems to be a better term for what modern scholars used
to call ‘transcendentals’ in the Middle Ages, since this is quite a late piece of terminology. For the
purpose of the present paper, the paraphrase ‘name which can be said of everything’ for the
group ens, unum, res, aliquid is probably the most precise; nevertheless I shall call such terms in
general ‘transcendentals’ since this denomination is more common among scholars.

300 L. Valente / Vivarium 45 (2007) 298-310

basic idea is that, in spite of the loss of the major part of the Aristotelian
corpus, the twelfth-century masters in logic still received some Aristotelian
theses concerning the notions of one and being via Porphyry and Boethius;
and that, on the basis of such theses, they were able to elaborate a sort of
proto-theory of the transcendental as trans-categorical terms.

2) The second aim of my paper is to show that this theory is centred on the
idea that there exists a particular group of names which have the specific
property that they can be said of everything; this group includes “being”,
“one”, “thing” and “something” (ens, unum, res, aliquid ), while “good”,
“true”, and “beautiful” are not part of the group. Twelfth century mas¬
ters in logic try to question the (originally Aristotelian) thesis that these
terms are equivocal, although they do not deny it completely.

Porphyry’s Isagoge and Its Tradition

In the section on species in his Isagoge , Porphyry first defines the notions of
most general genus ( generalissimum , following the Boethian translation), low¬
est species (specialissimum) , and of middle genus and species (ea quae in medio
sunt). Then he goes on to describe the difference between the tree of the genera
and any other tree, as for example the family tree: while in the family tree we
usually tend to go back to one ancestor, in the genus and species tree we are
compelled to stop at a set of ten ancestors, ten generalissima which are not spe¬
cies of a more general genus; these are the ten categories. Porphyry continues
writing that if anyone thinks that there is a unique most general genus to
which the ten categories belong as species, and that this most general genus is
being (ens), he is wrong. If being were the only genus, common to all things,
then all things would be called “beings” univocally. It is true that “being” can be
said of everything, but this happens only because “being” is then used equivo¬
cally: that is, by a nominal community, not by a community of definition. Being
then, for Porphyry, is not the most general genus, and the word “being” is used
equivocally when it is predicated of terms which fall under different categories;
or, at least, this is the thesis of Aristotle presented here by Porphyry. 6

6) Porphyry (1966), 11: 7-12: 6: “Sed in familiis quidem plerumque ad unum reducuntur princi-
pium, verbi gratia Iovem, in generibus autem et speciebus non se sic habet; neque enim est com¬
mune unum genus omnium ens, nec omnia eiusdem generis sunt secundum unum supremum
genus, quemadmodum dicit Aristoteles, sed sint posita (quemadmodum in Praedicamentis) prima
decern genera quasi prima decern principia. Vel, si omnia quis entia vocet, aequivoce (inquit) nun-
cupabit, non univoce; si enim unum esset commune omnium genus ens, univoce entia dicerentur,
cum autem decern sint prima, communio secundum nomen est solum, non etiam secundum

L. Valente / Vivarium 45 (2007) 298-310

301

In the Isagoge, Platonism and Aristotelianism are inextricable interlaced.
The thesis that “being” is said of everything non-univocally is present in
Plotinus, Enn. VI, 1.1., 15-30. Plotinus says that the Aristotelian categories
cannot include both the sensible and the intelligible worlds, since no term can
be said univocally of the sensible and the intelligible worlds. Plotinus is here
applying to the Platonic distinction between the sensible and the intelligible
worlds the following principle: there cannot be any synonymous genus for a
series of elements which are disposed on a scale of priority and posteriority—a
principle expounded by Aristotle in Metaphysics B 3, 999a6-10 and Eudemian
Ethics 8, 1218aIff. Moreover, Porphyry, in the Isagoge passage to which we
have referred, mentions Aristotle explicitly, and he may have had in mind not
only the Categories (ex. lal-15 and 3b7-9) and the Topics (109b4-7) concern¬
ing the notion of univocity and equivocity; but also Metaphysics B 3, 998b22
(where Aristotle explains that neither being nor one are genera even if they
“are said of everything which exists”), and T 2, 1003b5-6 (about the multiple
senses of “being”). Aristotle, then, may be not only the main direct source for
the theory of transcendental in the thirteenth century with his Metaphysics
and Ethics , but also one of its indirect sources, following a continuous line
which descends along the centuries and goes through the late ancient Platon¬
ism and Aristotelianism, that is, through Plotinus and the short treatise by
Porphyry; and mainly through Boethius.

Boethius's Commentary on Porphyry's Isagoge

In his second commentary on Porphyry’s Isagoge , commenting on the chapter
about species, Boethius confirms the theses about the term “being” expounded
by Porphyry that he had already noted in his first commentary:

1) There cannot be one unique principle for all things;

2) Some have asserted that a common genus for all things exists and that
this is being, since the term “being” is predicated of everything;

3) Aristotle denied that all things may be connected to one unique prin¬
ciple and he found ten most general genera for all things: these are the
ten categories;

4) When a term is predicated commonly of many things, it is not necessar¬
ily a genus for these things. A term which is said of many things is a
genus for those things only if it is said of them univocally, that is, with
the same definition; but this

definitionis rationem quae secundum nomen est.” Whether Porphyry assents here to the Aristote¬
lian thesis of the equivocity of the term “being”, or he is just referring it, is not clear from the text.

302

L. Valente / Vivarium 45 (2007) 298-310

5) is not true for the term “being”: “being” is said of all categories but not
with the same definition; therefore, it is said of all categories equivocally
and not univocally {id quod dicitur ens, etsi de omnibus dicitur praedica-
mentis, quoniam tamen nulla eius definitio inueniri potest quae omnibus
praedicamentispossit aptari, idcirco non dicitur uniuoce depraedicamentis,
id est ut genus, sed aequiuoce, id est ut uox plura significans) . 7 8 9

Boethius not only reaffirms what Porphyry had already said in the passage
from the Isagoge which he is commenting, but goes on to introduce the notion
of one along with the notion of being, probably having in mind Aristotles
Metaphysics (B 3, 998b22). Boethius says that being can not be genus of all
things since it is completely equivalent with one: both can be said of all things.
But it is not possible that the same species have two “coextensive” genera: it
follows that neither being nor one are genera of all things. Besides, being and
one cannot be genus or species of each other: in fact, we can neither say that
one is genus of being, nor that being is genus of one. They are so identical that
they cannot be predicated one of the other, but genus and species “do not
convert” {quod dicimus ens, unum est et quod unum dicitur, ens est; genus autem
et species sibi minime conuertuntur si igitur praedicatur ens de omnibus praedica¬
mentis, praedicatur etiam unum. [..] duo igitur aequalia singulorum praedica-
mentorum genera sunt, quod fieri non potest)}

It may be worth noting, as a sign of the great permeability between the
realms of logic and theology, that the principle of the convertibility of ens and
unum is explicitly asserted by Boethius in his theological treatise Contra
Eutychen et Nestorium. In the fourth chapter, Boethius deduces the necessity
for Christ to be one person from the thesis that a duplicity of persons would
make it impossible for Christ to be one being and so would make him to be
nothing at all, since “one and being convert” {Quod enim non est unum, nec esse
omninopotest; esse enim atque unum convertitur et quodcumque unum est est)} In
the twelfth century, Gilbert of Poitiers, commenting on this passage, does men¬
tion the principle of the convertibility of ens and unum as a “topic from which
philosophers derive many propositions” and a “proposition known perse ”. 10

7) Boethius (1906), 221: 13-223: 24. Cf. also Boethius (1880), 77: 1-13.

8) Boethius (1906), 223: 24-224: 17.

9) Boethius (2000), 220: 298-300.

10) Gilbert of Poitiers (1966), 300: 72-74: “Quod enim non est unum, nec esse omnino
potest: sicut etiam quod non est, non est unum. Est enim philosophis multarum proposi-
tionum locus et per se nota propositio quod esse atque unum uniuersaliter conuertitur.
Quodcumque enim est, unum est. Et quodcumque unum est, est. ”

L. Valente / Vivarium 45 (2007) 298-310

303

“Everything which is, is because it is one ”

Another passage from Porphyry’s Isagoge is interesting from our point of view.
After having declared that he will try to answer Porphyry’s question about the
status of genera and species, but before explaining his answer Boethius pres¬
ents an aporetic argument: one can neither say that genera and species are, nor
that they subsist only in the intellect. 11 While arguing the first part of the
aporia, that is, the thesis that genera and species are not, Boethius affirms that
genera and species, being common to more entities, do not possess unity; and
that therefore they cannot be absolutely, since “everything which is, is, because
it is one” ( omne enim quod est, idcirco est, quia unum est)) 2

We find here the prototype of the adage quidquid est (or omne quod est),
ideo est quia unum ( numero ) est: a formula which would become widely diffused
among the theologians of the 12th century: it is used for example in the com¬
mentaries on Boethius’ Opuscula by Thierry of Chartres and his school, 13 by
the Porretan anonymus Invisibilia Dei) A by Alain of Lille 15 and others. 16

Twelfth-Century Logic

In twelfth-century logic the transcendental terms appear mostly either within
a commentary on the Porphyrian passage about ens said equivocally of all
things, or within the discussion about the notion of equivocity. In both these
contexts these terms are characterised as those “names which can be said
of everything”. I shall now examine this theme in some individual twelfth-
century logicians. 17

The Dialectica ofGarlandus ofBesangon

In his Dialectica , Garlandus writes about the supercategoriality of being in the
first book, on non-complex terms, in the section on equivocity of the treatise

n) Boethius (1906), 161: 14-164: 2. Cfr. Maioli (1974), 132ff., with a translation of the
Boethian passage on pages 139f.; De Libera (1999), 205ff.

12) Boethius (1906), 162: 2-3.

13) Thierry of Chartres (1971), 80: 62. Cf. also William of Conches (1999), 187: 20-21.

,4) Invisibilia Dei (1973), 128: par. 39.

15) Alain of Lille (1981), 126: par. 3.

16) Cf. Vicaire (1937), passim, and Valente (2006).

17) For similar analyses in Introductiones Montane minores , Tractatus Anagnini, Ars Meliduna,
Dialectica Monacensis, Summe Metenses, and some fragments on grammatic, see Jacobi (2003).

304

L. Valente / Vivarium 45 (2007) 298-310

about categories. The Isagoge is clearly the text which Garlandus has in mind.
But Garlandus is not just a repetitor of the late antique tradition, he adds some
personal points to the subject. In particular, we can note the entry, along with
the term ens , of the term res : and we shall see this term appearing along with
ens more times in logic and in theology long before Albert the Great and
Thomas Aquinas would insert it in their lists of transcendentals in the wake of,
as it is usually said, the Metaphysics of Avicenna. 18 The main points of Garlan-
dus’s expositions are the following:

- ens and res are equivocal because of the will of the impositores (consilio ); they
have been chosen for the sake of brevity and they name briefly all things
together; 19

- ens is not the most general genus for all things since it is equivocal, because
it has different definitions, like the word canis m2Q

- ens is not the most general genus since, even if we assume the definition:
“being is what is perceived by the mind and distinguished by the reason”,
the term ens would still be equivocal; 21

- ens is not the most general genus for all categorical terms since every cate¬
gory has its own modi significandi\ but it cannot be admitted that the spe¬
cies of the same genus had different modi significandi. 12

18) On res as a transcendental in the 13th century see Ducharme (1940), Oeing-Hanhoff (1982),
Aertsen (2002), with bibliography; in this article Aertsen considers Avicennas Methapysics (in
part. tr. I, ch. 3) as the source for the introduction of the term res as a transcendental in medieval
philosophy. See also Ventimiglia (1998), on res as transcendental in connection with the theory
of the infinite names, and Hamesse (1982).

19) Garlandus (1939), 14: 10-13: “ Equivoca duobus modis dicuntur, casu et consilio. [...]
Consilio , idest per consilium nomine (siced.\ nomina?) imponentium, alia causa brevitatis, ut ens
et res, alia per similitudinem, ut homo pictus. Causa brevitatis ideo ut omnia insimul breviter
nominentur”. Res appears togheter with esse and unum as equivocal terms also in John of Salis¬
bury’ Metalogicon : John of Salisbury (1991), 177: 40-42: “Ceterum quoniam esse, et unum, et
res, multipliciter dicuntur, uidebit quisque quomodo interpretetur significationem”.

20) Garlandus (1959), 14: 14-17: “Contra illos qui disputant in Porphyrio dicentes ens esse
unum genus omnium, accipe hanc apertam rationem: ‘quare equivocum sit ens’, ideo scilicet
quia vadit ad diversa significata per diversas diffinitiones et significationes, sicuti ‘canis’
equivocum”.

21) Garlandus (1959), 14: 32-15: 4.

22) Garlandus (1959), 15: 5-19.

L. Valente / Vivarium 45 (2007) 298-310

305

Abelard’s Dialectica

In his Dialectica Abelard speaks of ens in relation with the discussion about the
different types of equivocity, and he seems to be trying to limit and circum¬
scribe its equivocity. It is worth noting that, as in the Logica Ingredientibus (see
below), here ens does not appear alone, but in a series of terms which now
contains res and “other names of the different categories”; names, Abelard says,
which are communia summa et suprema principia. They are equivocal because,
as the infinite names, they are not imposed on the basis of the fact that the
beings on which they are imposed have a common substance. Nevertheless
they are said of all things because they have a single imposition, and there is
one cause for this imposition: the cause for the imposition to all things of the
term res, ens etc. is to show that the things do exist:

Similarly both “non-body” and other infinite words are said to be equivocal to those things
which they nominate, since they are not imposed from a common substance. In this way
also Aristotle included with equivocals “ens” and “res” and other names which belong to
different categories, of which there can be no thing that is according to the same substantial
being—that is, they are general, overall names and supreme principles. For perhaps “ens”
and “res”, in that they have a single imposition, made once, on all things [...] also have a
single cause of imposition. For “ens” is imposed on all existing things because each is one of
the things that exist. 23

Moreover, ens , res and cetera diversorum predicamentorum nomina , if we con¬
sider carefully, do not have a definition, because we would be compelled to
insert in the definition the very same names which are to be defined. 24

23) Peter Abelard (1970), 565: 20-33: “Similiter et ‘ non corpus et cetera infinita ad ea que nomi-
nantur, quibus ex communi substantia non sunt imposita, dicuntur equivoca. Sic et 'ens et ‘ res
et cetera diversorum predicamentorum nomina, quorum nulla potest esse secundum idem sub¬
stantial esse—communia scilicet summa sunt et suprema principia—, equivocis Aristotiles
aggregavit. Nam et fortasse 'ens et ‘ res secundum id quod unam habuerunt ad omnia et semel
facta impositionem [...], et unam impositionis causam habere. 'Ens enim [in] omnibus exis-
tentibus inde est impositum quod aliqua sit rerum existentium.” Cf. also Theologia Christiana IV,
in Peter Abelard (1969a), II, 342-343, par. 157, where it is said that est and res are equivocal with
respect to the substance and the properties.

24) Peter Abelard (1970), 565: 36-566: 2: “Insuper, si diffinitionis proprietatem inspiciamus, 'ens
et 'res' omnino diffinitionem recusant. Si enim 'ens ita determines: ‘ res existens' , 'existens' , quod
idem est in sensu cum 'ens, bene ipsum non diffinit; idem enim per se ipsum minime diffiniri
potest. Si vero dicamus quod non est non-existens, ‘ non-existens' cognosci non potest nisi per
'existens'2

306

L. Valente / Vivarium 45 (2007) 298-310

Abelard's Logica Ingredientibus

In Abelards Logica Ingredientibus , Glossae super Porphyrium, where he discusses
the passage we are considering here from the Isagoge , we find a long discussion
about the equivocity of ens , and this discussion has its starting point as usual
in the second commentary on the Isagoge by Boethius. Concerning the fact
that ens is not a genus, Abelard just reports without any objections the Aristo¬
telian thesis which was presented by Porphyry. Then Abelard defines, follow¬
ing Porphyry, the ten categories as the ten names which are the origin of the
meaning of any other name. 23

After that, however, Abelard discusses the question of the equivocity of ens
with respect to the ten categories—and here his position seems to be different
from that of Boethius and Garlandus. Both Boethius and Garlandus had com¬
mented on Porphyry by explaining that the equivocity of ens with respect to
the categories had its origin in the fact that there is no common definition of
ens; therefore, ens has different meanings, exactly like canis. But Abelard inter¬
prets the Porphyrian formula “The commonness is just a nominal one, it
doesn’t entail an identity in definition” (“communio secundum nomen est
solum, non etiam secundum definitionis rationem”). It does not mean that ens
has different definitions, but rather that it does not have a common definition
in substantia with respect to the categories: this is a definition the terms of
which are said of the subject in the category of substance: the definitions of
animal and of rational being for example are definitions in substantia of man
or Socrates, while the definition of white thing is not one. 26

The same argument is presented in a passage from the Glossae super Prae-
dicamenta about the equivocal terms. Here Abelard speaks about the equivocity
not only of the term ens, but also of the term unum and of “every name which
contains things of different categories.” Such terms are not equivocal in the
sense that they have different meanings, like canis , but in the sense that they

25) Peter Abelard (1919-1933) ,51:19-52: 27. Cf. also Abelard, Editio supra Porphyrium , in Abelard
(1969b), 17: 27-18: 19.

26) Peter Abelard (1919-33), 53: 5-31; in part. 16-23: “Cum itaque omne genus de specie vel
individuo in substantia praedicatur, ita videlicet, quod suus sensus in illis concludatur, substantia
vero vel qualitas nil de sententia entis denotat, ut supra diximus, non potest ulla definitio data
secundum nomen entis in substantia de praedicamentis praedicari. Unde entis nomen Aristo-
teles ad praedicamenta univocum non appellat, sicut nec album univoce de subiectis substantiis
praedicari vult, non quia habeat diversas definitiones, sed quia non habet communem substan¬
tiae rationem ad primas substantias.” On the “definitio substantiae”, cf. 118: 4-7. On the
difference between Abelard and Boethius on this point see Marenbon (1992).

L Valente / Vivarium 45 (2007) 298-310

307

do not have a definition which may be predicated of the predicaments in the
category of substance. 27

With Abelard, then, we find some new elements which go in the direction
of the future theory of transcendental: the list of such terms increases and
contains now, beside ens and unum , also the term res as well “all names which
are said of all categories”; the assertion of their equivocity is questioned and it
is said that there is in any case a common cause of the imposition of such
names to all things.

The Anonymus D’Orvillensis and Anonymus Patavinus on the Categories
of Aristotle

The Aristotelian and Porphyrian thesis of the equivocity of ens y unum , and
similar general terms is also questioned in two twelfth-century commentaries
on the Categories', the Anonymus DOrvillensis 1 * and the Anonymus PatavinusP
These two commentaries agree with Abelard’s observation that the transcen¬
dental names cannot be defined. This is why—these two texts seem to say—
such terms are in a middle position between equivocal and univocal terms.

Ens , says the Anonymus D’Orvillensis , is both equivocal and univocal with
respect to every thing:

We concede that there even is a signification in which all equivocals are univocals, for all share
the name ens in the same signification, and thus <they share it> univocally; and they all share
the same name, but no account assigned according to the name, and thus they share it
equivocally; thus all things are <both> equivocal and univocal under this name ens. 30

A similar argument is expressed in the other commentary on the Categories ,
the Anonymus Patavinus. The author, commenting on the section on the

27) Peter Abelard (1919-33), 117:36-118: 1: “Quo etiam modo ens et unum et quodlibet nomen
res diversorum praedicamentorum continens de ipsis praedicamentis aequivoce dicitur, non
quod multa sint vocabula in sensu, sicut est canis, sed quia definitionem non habent in substan¬
tia de praedicamentis praedicabilem.”

28) Ms. Oxford, Bodleian Library, D’Orville 207, ff. lra-24vb, ed. Ebbesen (1999); cf. Ebbesen
(2000) and Spruyt (2003).

29) Ms. Padova, BU 2087, ff. lra-48vb. On this commentary see De Rijk (1966) and Spruyt
(2003).

30) Ed. Ebbesen (1999), 239: “Item, queritur utrum aliqua aequivoca sint univoca. Quod con-
cedimus, immo in aliqua significatione omnia aequivoca sunt univoca, omnia enim participant
hoc nomine ‘ens’ in eadem significatione, et ita univoce; et omnia participant eodem nomine et
nulla ratione assignata secundum illud, et ita aequivoce; ergo omnia aequivocantur et univocan-
tur sub hoc nomine ‘ens’.” Translation in Ebbesen (2003a), 358f.

308

L. Valente / Vivarium 45 (2007) 298-310

equivocal terms of the Categories , first of all describes the terms res , ens y unum ,
and aliquid as, among the equivocal terms, those which are more noteworthy
than others; moreover, Aristotle would affirm that they are equally predicated
of everything. 31

We find here, together with the terms ens, unum , and res , which we have
already seen, also aliquid. This is an important integration, especially if we
consider that this term plays an important role in twelfth-century discussions
about theological language 32 as well as in the mature theories of transcenden¬
tal, for example in Thomas Aquinas. 33

The Anonymous then poses the question whether the Aristotelian descrip¬
tion of equivocals is valid for every equivocal term, that is also for res , ens and
unum. In the argument for a positive answer, the Anonymous explicitly men¬
tions the passage from the Isagoge from which we have started; the argument
for a negative answer is the thesis, which we have seen in Abelard, that these
terms can not be equivocal on the basis of different definitions, since they do
not have a definition at all: there is no more general term which can be used
in order to ‘begin their definition. 34

The solution to the question is not clear: the author nevertheless seems to
favour the thesis that they are equivocal.

31) Anonymus Patavinus, Comm. Cat. [C 15], Padova, BU, 2087, f. 2va: “Iterum, prius tractat
de aequivocis quam de univocis quia inter aequivocos comp... dicuntur ( transcr. Spruyf. com-
muniter dantur; comprehenduntur? computantur?) termini digniores aliis sicut ‘res’ et ens’
‘unum’ et ‘aliquid’, de quibus dicit Aristoteles quod aequaliter de omnibus rebus praedicantur, et
ex hac dignitate prius tractat de aequivocis quam de univocis.” I would like to thank Sten
Ebbesen for allowing me to use his transcritption of passages from the Anonymus Patavinus. The
same passage is quoted, but with a different reading, in Spruyt (2003), 333, n. 14.

32) C f. Valente (2005), 223-237.

33) On aliquid as a transcendental see Schmitz (1981), Ventimiglia (1997), 208fL, Ventimiglia
(1998), Rosemann (1998). Res and aliquid are considered as transcendental not only by Thomas
Aquinas (De veritate q. 1 a. 1 .et alibi), but also by Bonaventure, Duns Scoto, the De natura generis
falsely attributed to Thomas Aquinas and, in logic, by the Dialectica Monacensis (res) and the
SummaeMetenses (aliquid). Cf. De Libera (1994), 142 and l48ff.

34) Ms. Padova, BU 2087, f. 2va; passage quoted by Ebbesen (2003a), 359, n. 44: “Quaeritur/
Quare [[i]] (?) haec descriptio (sc.: Aequivoca dicuntur quorum nomen commune est, ratio vero
assignata secundum id nomen est diversa) conveniat omnibus aequivocis. Si omnibus, ergo istis,
sc. rei, enti et uni. Quod haec si<n>t aequivoca probatur ex tpositionif Ar(istotelis) et Por(phyrio)
dicente ‘Nam si quis vocet omnia entia, aequivoce, inquit, nuncupabit et non univoce.’ Et si
hoc nomen ‘ens’ convenit omnibus aequivoce et non univoce, et sic est aequivocum. Sed
probatur quod non est aequivocum, quia non est sic commune pluribus quod ratio assignata
secundum illud nomen sit diversa, nulla enim ratio eius est. Quod nulla sit probatur: Omnis

L. Valente / Vivarium 45 (2007) 298-310

309

Ars Meliduna

Another, important logical treatise from the twelfth century, the Ars Meli-
duna , 35 affirms that the terms—in this order— res y aliquid , ens, unum “suit
everything in accordance with their denotation” ( conveniunt cuilibet rei per
appellationem ), and that they do not signify a universal. If they did, we would
have a unique genus for everything, but Aristotle explicitly denies this. More¬
over, these terms—as Abelard also says—do not show what the thing of which
they are predicated is, but simply that this thing exists. 36 This text—which has
already been analysed by Jacobi 37 —seems to me very significant also because
it on the one hand evidently refers to the Isagoge in asserting that ens is not the
most general genus, and on the other hand anticipates a theme which will be
important in the future doctrine of transcendentals. It is said, in fact, that res,
ens, unum , and aliquid “conveniunt cuilibet rei per appellationem' : it is hard
not to see here an anticipation of the thesis, generally accepted in the thir¬
teenth century, that the transcendental terms transmit different meanings, but
have same reference (here appellatio , but later it will be suppositio).

Compendium logicae Porretanum

That the terms ens, unum, res and aliquid do not signify any universal is
affirmed also in an important treatise on logic written by a disciple of Gilbert
of Poitiers, the Compendium logicae Porretanum. In this text the list of the four
terms appears twice. The first time, in the first part dedicated to problems of
“philosophical grammar”, in a paragraph devoted to the verb “be” {sum). Here
it is said that ens, unum, aliquid do not signify either a universal or a singular,
but that they have an unclear meaning, or that they have no meaning at all.

enim definitio incipit a maiori termino et descendit ad aequalia, ut sit eius aequalis; sed nullus
terminus incipit (!) definitionem eius, quia nullus terminus est maior hoc termino ente, quia
nullus terminus de pluribus dicitur quam iste, quia hie terminus de omnibus praedicatur et sic
nullus terminus incipit eius definitionem, et sic nulla est eius oratio, quia illud nomen non vide-
tur esse aequivocum.”

35) Contained in the ms. Oxford, Digby 174, ff. 21 lra-24lrb. The treatise can be dated between
1154 and 1180. Cfr. De Rijk (1962-67) II i, 280-281.

36) Ms. Oxford, Digby 174, f. 221va. Cf. De Rijk (1962-67) II i, p. 309: “Nominum alia conveniunt
cuilibet rei per appellationem, alia non. Nullum nomen conveniens cuilibet rei significat universale ut
res, aliquid, ens et unum. Genus uero quod magis videre<n>tur significare <significare> non possunt
ne sit unum communem genus omnium, quod prohibet Aristoteles. Preterea assignatione sua non
ostenditur de aliquo quid ipse sit sed simpliciter quod sit.”

37) Jacobi (2003), 26ff.

310

L. Valente / Vivarium 45 (2007) 298-310

The group ens y unum , and aliquid is connected with the grammatical prob¬
lematic of the classification of verbs as adjectival verbs, substantival verbs, and
vocative verbs. Substantive and vocative verbs {sum and vocor) have no clear
meaning, just as ens, unum and aliquid do not. The author seems to say—but
the text is unclear to me—that as sum doesn’t signify either a universal or a
singular property, in the same way ens , which derives from sum , has an unclear
meaning; and this holds also for unum and aliquid since they are convertible
with ens and with each other. 38 The second occurrence of the list of the
transcendental terms in the Compendium is in the third part of the work,
dedicated to problems of ontology, within the reflection about singulars, indi¬
viduals and universal. In Porretan ontology, everything is numerically one
(singular); therefore, the terms singulare\ unum , ens , and aliquid range over the
same objects {paria sunt ). 39

Conclusion

In Latin logic, before the entry of the new Aristotle and Avicenna in the thir¬
teenth century, we can find a clear line of thought which—treating of the
“names which can be said of everything”—goes in the direction of the later
doctrine of transcendental. A particularly important role in this respect was
played by the reception of Porphyry’s Isagoge and of the commentary by
Boethius on it. Using these texts, medieval masters elaborated in a continuous
line over the centuries old ideas and principles which originally derive from
the Amtiquity and the Late Antiquity. By the end of the twelfth century, logical
reflection about the “overcategorical” names enters the discussions about the
univocity or equivocity of theological language. But this is another story. 40

38) Compendium logicae Porretanum (1983), 10: “[...] nec universale nec singulare est quod
significat hoc verbum ‘sum’. [...] At cum ‘ens’ retineat significationem sui verbi, ut finitam
habeat significationem, ad quod censequuntur et ‘unum’ et ‘aliquid’ et ad idem convertantur
utrumque, confusionis patitur detrimentum, nec reperitur ad quod significandum aliquod
eorum fuerit inventum, quod confirmat Aristoteles in Analeticis: ‘Non sunt sumenda pro mediis
quecumque do omnibus dicuntur’.”

39) Compendium logicae Porretanum (1983), 48: “Notandum quod singulare dicitur res discreta
numero, sed quod paria sunt unum et ens et aliquod singulare et aliquid”.

40) Cf. Valente (2003). For the use of the term suprapraedicamentale see Summa Colligite frag-
menta , Ms. Miinchen Clm 28 799, cit. in Heinzmann (1974), 179, n. 21. The Summa Breves dies
hominis defines unum , ens and aliquid as “those names which transcend the most general genera”:
“ilia quae transcendunt generalissima”, and asserts that at least these names are said of God
properly (Ms. Bamberg, Staatsbibliothek, Patr. 136, f. 2rb).

BRILL

Vivarium 45 (2007)311-327

VIVA

RIUM

www.brill.nl/viv

Metaphor and the Logicians from
Aristotle to Cajetan

E. Jennifer Ashworth

University of Waterloo, Canada

Abstract

I examine the treatment of metaphor by medieval logicians and how it stemmed from
their reception of classical texts in logic, grammar, and rhetoric. I consider the relation
of the word ‘metaphor to the notions of translatio and transumptio , and show that it is
not always synonymous with these. I also show that in the context of commentaries on
the Sophistical Refutations metaphor was subsumed under equivocation. In turn, it
was linked with the notion of analogy not so much in the Greek sense of a similarity
between two proportions or relations as in the new medieval sense of being said secun¬
dum prius et posterius. Whether or not analogy could be reduced to metaphor, or the
reverse, depended on the controversial issue of the number of acts of imposition
needed to produce an equivocal term. A spectrum of views is canvassed, including
those found in the logic commentaries of John Duns Scotus.

Keywords

analogy, equivocation, imposition, metaphor, Sophistical Refutations commentaries

In this paper I shall sketch an answer to a series of questions about the treat¬
ment of metaphor by medieval logicians. One question is linguistic: are the
words “translatio” and “transumptio” synonyms of the word “metaphora”?
Three other questions concern analogy and equivocation. First, is metaphor a
type of equivocation? Second, is metaphor a type of analogy and if so, what
type? Is it linked with analogy in the Greek sense of a similarity between two
proportions or relations, or with analogy in the new medieval sense of being
said secundum prius et posterius because of some attribution? Third, how many
acts of imposition are required for the production of analogical terms and
metaphors? This last issue is particularly important, given that words are said

DOI: 10.1163/156853407X217795

312

E. J. Ashworth / Vivarium 45 (2007) 311-327

to be used proprie only when used in accordance with an act of imposition,
and that metaphors are normally said to be taken improprie. I will take up
these questions in the context of three sets of texts. I will start with some
remarks about the texts of Aristotle and their reception in the Middle Ages.
Secondly, I will look at translatio and transumptio in ancient grammar and
rhetoric. Finally, I will look at medieval logic texts, especially commentaries
on the Sophistical Refutations.

My study will show how ancient traditions in logic, grammar and rhetoric
were interwoven and used to tackle specifically medieval problems. Aristotle
played a prominent role in the story, but not primarily because of his explicit
discussions of metaphor in his Poetics and Rhetoric. Stoic thinkers contributed
the theory of tropes or figures of speech; and Neoplatonic commentators such
as Porphyry influenced Boethius’s discussion of equivocation and metaphor.
The thirteenth century theory of analogy itself grew out of the interweaving
of problems in Christian theology, Aristotelian metaphysics and Aristotelian
logic, but was enriched by the long Greek and Arabic tradition of analysing
ambiguous terms as being said secundum prius etposterius. The resulting syn¬
theses, especially in late thirteenth and early fourteenth century British logi¬
cians, show a skilful use of whatever parts of ancient traditions seemed relevant
to the particular interests and doctrines of the author in question.

Aristotle

Aristotle discussed metaphor in his Rhetoric and his Poetics. These two works
came to be regarded by the late Greek commentators of the School of Alexan¬
dria as part of the Organon, that is, as part of logic in a wide sense, and the
Arabs followed this established tradition. Albert the Great, Aquinas, and Simon
of Faversham, among others, followed the Arabs, at least in principle. However,
the practice was rather different, given the relative inaccessibility of the two
works. It is tempting to think that they contain the origins of medieval discus¬
sions of metaphor, but at least so far as direct influences go, this hypothesis
cannot be sustained.

Aristode gave his standard definition of metaphor in Poetics I457b7-18, writing:

Metaphor consists in giving the thing a name that belongs to something else; the transfer¬
ence being either from genus to species, or from species to genus, or from species to species,
or on grounds of analogy.... That from analogy is possible whenever there are four terms so

E. ]. Ashworth / Vivarium 45 (2007) 311-327

313

related that the second is to the first, as the fourth is to the third; for one may then put the
fourth in place of the second, and the second in place of the fourth. 1

We should note that two of these types of metaphor, from genus to species and
from species to genus were later classified by Quintilian as types of synecdo¬
che. 2 In accordance with his extensional definition, synecdoche differs from
metaphor because in synecdoche one moves between more and less in things
of the same nature, whereas in metaphor reference is made to a different type
of thing. 3

William of Moerbeke translated the Poetics in 1278. However, only two
manuscripts survive, and for all practical purposes the translation remained
unknown. On the other hand, in 1236 Hermannus Alemannus translated the
middle commentary of Averroes on the Poetics . 4 This translation was reason¬
ably well known, and it gave rise to at least one commentary by Bartholomew
of Bruges, a short quaestio , and some glosses. 5 In some places (e.g. 59-60)
Hermannus does use the words “metaphora”, “metaphorica” and “meta-
phorice”, but when he refers to the standard Latin examples of metaphor and
amphiboly, “pratum ridet” and “litus aratur”, he speaks rather of “translatio”
and “transumptio” (42). The latter is associated with proportionality. The
work gives no very clear idea of the definition of metaphor, and the four types
of metaphor are included, without emphasis, in a discussion of various ways
in which words may be used, including transumption or transmutation to an
extraneous use (67). The example of analogy (without any use of that word) is
presented as a change of proportions, as when old age is called the evening of
life and evening is called the old age of the day on the grounds that the relation
( proportio ) of old age to life is much like {quasi) the relation of evening to
the day.

Aristotle’s Rhetoric was better known. There were at least two translations
dating from the middle of the thirteenth century, followed by that of William
of Moerbeke, finished around 1269 and transmitted by a large number of
manuscripts. In this work, Aristotle associates metaphor and simile, writing

n Quotations are taken from Aristotle (1984). However, I have modified them where necessary
to conform with the medieval Latin translations.

2) Quintilian (1970), 8.6.19, 465-466. For discussion, see Meyer (1993), 9.

3) See Grondeux (2002), 121.

4) See AL XXXIII.

5) See Dahan (1980).

314

£ J. Ashworth / Vivarium 45 (2007) 311-327

(l406b20-24): “The simile also is a metaphor; the difference is but slight.
When the poet says:

He leapt on the foe as a lion,

this is a simile; when he says of him ‘the lion leapt’, it is a metaphor—here,
since both are courageous, he has transferred to Achilles the name of‘lion’.”
Aristotle also writes (141 lal-3): “Of the four kinds of metaphor the most tak¬
ing is the proportional kind. Thus Pericles, for instance, said that the vanish¬
ing from their country of the young men who had fallen in the war was ‘as if
the spring were taken out of the year’.” However, in this passage he does not
explain what the other three kinds of metaphor are.

At least four commentaries on Aristotle’s Rhetoric still exist, though both
surviving manuscripts of John Buridan’s Quaestiones lack the material on met¬
aphor. 6 I shall focus on the popular commentary by Giles of Rome, written
between 1271 and 1274. Giles follows Aristotle’s text very closely, though not
always with understanding, apparently owing to errors in his copy of Moer-
beke’s translation. For instance, when Aristotle (141 Ial3-14) speaks of Sestos
as “the treasure trove of the Peiraeus”, 7 Giles translates “the man called Sextus
was a robber of the pyre, that is, of fire” 8 which means, Giles tells us, that he
was very audacious. According to Giles (f. 103 rb), the four kinds of metaphor
to which Aristotle refers are the following: asteyum , also called asteycum (e.g.
f. 103 rb), proverbium, transumptio , and assumptio or assimilatio. Transumptio
is the simplest kind of metaphor, as when we say of Achilles that he is a lion or
of the young men that they are spring-like.The other types all add something
to the basic metaphor (f. 105 rb-va). Asteyum is the result of a problem with
Moerbeke’s text. Moerbeke had left the word “asteios”, which roughly means
lively or witty, in Greek. Giles took asteyum to be a kind of metaphor to which
is added a teaching function, and he called it disciplinativa. It also has a link
with analogy, because the strongest kind adds the teaching function to a met¬
aphor involving analogy (f. 104 va). Proverbium has to do with proverbs, and
Giles explains that this is the kind of metaphor to which is added the element
of a common usage. Assimilatio , the word that William of Moerbeke had used
to translate simile, is the kind of metaphor to which is added a certain analogy

6) Fredborg (1976a), 47.

7) AL XXXI, 1-2, 299: 27-8, ‘Sestum autem predam Pireei.’

8) Giles of Rome (1968), f. 103 va, ‘Sextus, id est quod ille homo, erat praedo pyretri, id
est, ignis.’

E. J. Ashworth / Vivarium 45 (2007) 311-327

315

(analogiam quandam ), for example when one says that the young men are to
the city as the spring-time is to the year. We should note, however, that in this
case it is the explicit analogy in the Greek sense that is a type of metaphor,
whereas for other logicians an underlying analogy was what explained the
metaphor actually used. Thus Thomas de Wyk and Walter Burley both claim
that we say “foot” of a bridge because the bottom of the bridge supports the
bridge in the way that a man is supported by his foot. 9

Apart from the Rhetoric and Poetics , Aristotle speaks of metaphor in the
Posterior Analytics, in a passage (97b29-39) where he warns against the use of
metaphors and metaphorical expressions in definitions. In the same place he
makes a passing reference to the word “sharp” (acutuni), which he had described
as a metaphor in De anima 420a29, and, in Topics 1.15 (107a 14-18), as equiv¬
ocal when said of a vocal sound (vox), an angle and a knife. In commentaries
on the Posterior Analytics, notably those by Giles of Rome and Paul of Venice,
these remarks led to some discussion of the way in which “acutum” could
be both equivocal and a metaphor. Giles of Rome explains that many cases
of equivocity are caused by those metaphors which involve transumptio . 10
Because a sharp object penetrates magnitudes rapidly, the term is transferred
to those tastes that penetrate the taste buds, and then to those vocal sounds
that penetrate the hearing. He added that when Aristotle said that equivoca¬
tion was worse than metaphor, he was referring to pure equivocation, not the
kind that is equivalent to metaphor and transumptio, and which is based on
some similitude. Paul of Venice echoes Giles’s remarks in his own commentary
on the Posterior Analytics, though he does not refer to transumptio . n

Aristotle gave similar warnings against the use of metaphor in definitions
in Topics 6.2 (139b33-l40a2). He also remarks there (I40a9-ll) that all
metaphors are based on some kind of similitude, and earlier, in Topics 1.17
(108a6-l6) and 18 (108b24-27), he had explicitly linked similitude with
what became standard examples of analogy in the Greek sense. One is princi-
pium said of a point on a line and the unit in number, and the other is sight,
which is to the eye as intellect is to the soul. The latter example was also
influential through its use in Nicomachean Ethics 1.6 (1096b29-30). The pas¬
sages in the Topics introduce us to the next notion to be examined, translatio.

9) Thomas de Wyk, Fallaciae , in Ebbesen (1998), 142-143; Burley, Super tractatum fallaciarum
in Ebbesen (2003c), 202.

10) Giles of Rome (1967a), sig. p 6 ra-rb.

M) Paul of Venice (1976), sig. x 5 vb-6 ra.

316

E. J. Ashworth / Vivarium 45 (2007) 311-327

for in Boethius’s translation of the Topics , the Greek metaphora is always ren¬
dered as translatio.

Translatio

Translatio is part of the theory of tropes or figures of speech, which is Stoic in
origin. 12 According to Quintilian, “A trope is the change ( mutatio ) of a word
or phrase from its proper signification to another signification, with some
<added> force ( cum uirtute)L x}> Donatus, the fourth century grammarian
whose discussion of tropes in Book 3 of his Ars Maior became standard for
medieval undergraduate instruction, wrote: “A trope is an expression trans¬
lated ( translata ) from its proper signification to an improper similitude (ad
non propriam similitudinem) , either for ornament or out of necessity.” 14 Dona¬
tus s wording suggests a general sense of translatio , but for Quintilian translatio
(or tralatio in the variant found in the critical edition) is the trope that corre¬
sponds to the Greek metaphora. It is “the transfer of a noun or verb from the
place in which it is proper to another place where the proper word is lacking
or where the word which has been transferred is better.” 15 The earlier Rhetorica
ad Herennium had added a reference to similitude as justifying the transfer,
and had noted that the reasons for translatio included brevity, ornament, and
the avoidance of obscenity. 16 Quintilian listed four types of metaphor: the
transference from an animate being to another animate being, from an
inanimate being to another inanimate being, from an animate being to an
inanimate being, and from an inanimate being to an animate being. 17 This
fourfold division, which is found in Donatus, and in other authors, such as
Bede in his Liber de Schematibus et Tropis , 18 seems to be an elaboration of
Aristotle’s type of metaphor from species to species, and no mention is made
of analogy in the Greek sense.

12) For discussion and further references, see Meyer (1993), 10 and Holtz (1981), 200.

,3) Quintilian (1970), 8.6.1, 462: 11-12.

14) Donatus, Ars Maior \\\.6 in Holtz (1981), 667: 2-3.

15) Quintilian (1970), 8.6.3, 463: 3-5. Later Quintilian (1970), 8.6.34-5, 469 says that cata-
chresis or abusio gives a name to things that have no name, and that it differs from metaphor
because in metaphor there was a different word {abusio est ubi nomen defuit, tralatio ubi aliud

fait).

16) Anonymous (1964) IV (V) 45, 157: 27-158: 1; V (V) 34, 158: 5-16.

17) Quintilian (1970), 8.6.9-10, 463-464.

,8) Donatus III, 6, in Holtz (1981), 667: 6-7; Bede (1863), 611: 24-26.

E. J. Ashworth / Vivarium 45 (2007) 311-327

317

Translatio was linked with equivocation in the Categories commentary of
Boethius. He follows Greek commentators, notably Porphyry, in making a
prior distinction between chance equivocals which, according to later logi¬
cians, must involve two separate impositions of the one word, and deliberate
equivocals, where the different senses are linked in some way. There are four
kinds of deliberate equivocation. Similitude, in the sense of a similarity of
external form, exemplified by a true man and a painted man; analogy in the
Greek sense, exemplified by principium said of unity with respect to number
and point with respect to a line; “of one origin”, exemplified by medicinale\
and “in relation to one end”, exemplified by salutaris or, for later authors,
sanum , said of animals, their diet, and urine. 19 From the 1240s onward, the
last two subdivisions were classified as analogy in the new sense of being said
of two things in a prior and a posterior way ( secundum prius etposterius ), and
often an extra subdivision, relation to one subject, was added to accommodate
the example of ens. 20 Boethius then goes on to say that there seems to be
another mode of equivocation that Aristotle is silent about, namely transla¬
tion He cites the example of “pes”, said of a ship and of a mountain. He states
that translatio has no property of its own ( translatio nulliusproprietatis est ), by
which he means that it is not a particular class of word. It may or may not fall
under equivocation. If, for reasons of ornament, one calls a steersman a chari¬
oteer, the word “charioteer” is not equivocal, for the man is already properly
named steersman. However, if the object has no name of its own, as in the case
of the picture of a man, one can transfer the word “homo” from the living man
to his picture, and so the word “homo” is equivocal. Note that we are encoun¬
tering homopictus for the second time, for he was also the standard example of
similitudo , the first subdivision of deliberate equivocation.

The distinction between two types of translatio , found in Boethius and in
the rhetoricians, 22 as well as in Petrus Helias, 23 seems to cause a problem, given
that metaphor seems to involve an ephemeral, ad hoc use of words, related to
a particular language and to particular circumstances. Some commentators on
the Sophistical Refutations explained that it was because of this specificity that

19) Boethius (PL 64, 166 B-C). Cf. Simplicius (1971), 42: 62-43: 87.

20) See Ashworth (1992, 123). For secundum prius et posterius in Arabic sources, see Wolfson
(1973).

21) Boethius (PL 64, 166 D-167 A). Anonymus Parisensis, in Ebbesen and Iwakuma (1990), 74
reports fully on Boethius’s discussion. Cf. Fallacie Parvipontane in De Rijk (1962-1967), I, 334-333.

22) Cf. Quintilian (1970), 8.6.6-8, 463 on necessity and ornament.

23) Petrus Helias (1993), 216: 80-81.

318

E. J. Ashworth / Vivarium 45 (2007) 311-327

Aristotle, when he listed the three modes common to equivocation and
amphiboly, gave no example of the second mode which they identified with
metaphor. 24 This characterization of metaphor fits well with the idea of using
another word for something that already has its own name, such as “laughing”
instead of “flowering”, for the sake of ornamentation. It does not fit so well
with the idea of using a word with an established sense to name something
that has no name of its own, for that seems to imply a certain permanence. Of
the logicians I have read, only Peter of Spain notes that metaphors can become
fixed in the language, and thereby turn into straightforward equivocal terms. 25
We should note that in medieval practice fixed metaphors include those pri¬
marily introduced for the sake of ornamentation, such as the laughing mead¬
ows which provided a stock example from the twelfth to the sixteenth century,
and that these meadows became not just a stock example of metaphor but of
equivocation. 26

Another classical source that brought metaphor, under that name, together
with equivocation is Simplicius’s commentary on the Categories , translated by
William of Moerbeke in 1266. Simplicius remarked (43) that there was a pos¬
sible link beween one of the subdivisions of deliberate equivocation, analogy
in the Greek sense, and metaphor, when he wrote “Others, among whom is
Atticus, bring together the mode according to metaphor (secundam meta-
phoram) and the mode according to analogy (secundum analogiam) and affirm
that their reunion constitutes one single mode of equivocation.” He went on
to quote and elaborate on Porphyry’s point that when a word is transferred to
something that has a name, there is metaphor and no equivocation, but when
it is transferred to something that has no name, it is not a case of metaphor at
all, but is straightforward equivocation. The example cited is that of “foot”,
said on the one hand of the lower part of a mountain in place of the Greek
word “hyporia” and on the other hand of a table or bed, because of the simili¬
tude to the foot of an animal. Here there are two rationes and one common
name. Simplicius added (43-4) that the first case will count as equivocation if
similitude is involved.

24) Giles of Rome (1967b), f. 11 va-vb; Anonymus Monacensis, Commentarium in Sophisticos
Elenchos , in Ebbesen (1997), 145; Nicholas of Paris, Notulae super librum Elenchorum, in Ebbesen
(1997), 174.

25) Peter of Spain (1972), 113: 12-26.

26) For Abelard in Glosae super Praedicamenta , prata rident’ was not a case of equivocation: see
Rosier-Catach (1997), 156-157.

E. J. Ashworth / Vivarium 45 (2007) 311-327

319

To return to the word translation we cannot assume that even in the late
classical period it was always equivalent to metaphor as one particular figure of
speech [cf. Donatus, above]. In De Doctrina Christiana (II, 10, 15) Augustine
divides signs into two groups, propria and translata. Proper signs are those
used to signify those things they were instituted to signify (propter quas sunt
instituta) and transferred signs seem to be those used in accordance with any
figure of speech. He gives the example of “ox” used in “You shall not muzzle
an ox when it is treading out the grain” to stand either for an actual animal or
for an evangelist. In general, he argues for the importance of knowing the
tropes in order to dissolve ambiguities (III, 30, 41), and he makes specific
reference to metaphor and catachresis (III, 29, 40). This generality is even
clearer in Augustine’s De Dialectica (112-16) where translatio is introduced as
a type of equivocation depending on the use of words (ex loquendi usu). The
first case cited is that of similitude, as when a man and his statue are both
called “Tullius”. The second and third cases involve parts and wholes, the
fourth and fifth involve species and genus, the sixth and seventh involve cause
and effect, and the eighth and ninth involve the contained and the contained.
In other words, they cover synecdoche and metonymy. Another influential
text was De trinitate I, 1,2, where Augustine remarks that those things that are
said proprie of God and are not found in creatures, are rarely affirmed in Scrip¬
ture, and cites Exodus 3.14 as an exception. This remark was one of the inspi¬
rations for a prominent movement in twelfth century theology which held
that all words, with the possible exception of “Qui est”, are said translative
of God. In my view, it is clear that Gilbert of Poitiers and Alan of Lille,
to mention just two names, were using translatio in a much wider sense
than mere metaphor. 27 Similar extensions are found in the logicians, and Irene
Rosier-Catach (1999a) has recently studied the wide use of translatio in Peter
Abelard’s logical writings.

Transumptio

A word with a similarly wide use, and that often replaced translation was tran-
sumptio (or transsumptio ). For Quintilian, transumptio was metalepsis, which
occurs when there is a double transference, that is to say, when the word in
question replaces another word, which in turn replaces another word and so

27) See Ashworth (forthcoming), Valente (in press), Rosier (1995a), 135-168, Rosier-Catach
(1997).

320

E. J. Ashworth / Vivarium 45 (2007) 311-327

on. For instance, we might replace the word cano (I sing), by canto (I chant),
and then replace canto by dico (I say) in order to say that I say instead of saying
that I sing. 28 There are elaborate discussions of transumptio in later medieval
writings on the poetic arts, 29 but for most logicians the word transumptio was
merely a synonym of the word translation and enjoyed the same ambiguities.
For instance, in the thirteenth century we find Roger Bacon, who was strongly
influenced by Augustine’s De dialectica , using the words transsumptio and
transsumere in his De signis to cover a variety of cases of equivocation and
analogy. 30 In his Compendium Studii Theologiae (116) he notes that figurative
language, as in Pratum ridet , constitutes an improper mode of equivocation,
so it is unlikely that he meant transumptio to be equivalent to metaphor.

We find the wider sense in various logical sources, at least three of which
bring together Boethius’s four types of deliberate equivocation with the notion
of transumptio. Anonymus Aurelianensis, writing perhaps between 1160 and
1180, 31 said that the second type of equivocation was when a name signified
one thing principally, but was transferred to signify another thing either from
proportion or from similitude or from some other cause, and he gave the
examples of homo pictus and the laughing meadows. 32 There is a clear hint
of Boethius’s four types of deliberate equivocation here, as there is in the
thirteenth century Categories commentary of Robert Kilwardby. 33 It should
also be noted that Anonymus Aurelianensis remarks that the first case of
equivocation, Boethius’s chance equivocation, involves diverse institutions,
which suggests that the other types he lists do not. In the late thirteenth
century, Burley explicitly identifies types of transumptio with Boethius’s types
of deliberate equivocation, and he insists that there is just one act of imposi¬
tion involved. 34

In discussions of supposition theory, the word transumptio also has a wider
sense. For instance, Walter Burley explains that supposition is improper when
a term supposits because of transumptio or the use of words (ex transsumptione
vel ex usu loquendi ), and Vincent Ferrer mentions this view. They both go
on to cite antonomasis, synecdoche, and metonomy as producing improper

28) Quintilian (1970), 8.6.37-39, 470.

29) See Purcell (1987).

30) Fredborg, Nielsen and Pinborg (1978), e.g. 109-115, 128-130.

30 Ebbesen (1979), xxviii.

32) Anonymus Aurelianensis, Commentarium in Sophisticos Elenchos , in Ebbesen (1979), 86.

33) Kilwardby, in Lewry (1978), 374.

34) Burley, Quaestiones super Sophisticos Elenchos, in Ebbesen (2005c), 281.

£ J. Ashworth / Vivarium 45 (2007) 311-327

321

supposition. 35 We find similar material in the Logica “Cum sit nostra ”, Roger
Bacon, William Ockham and John Buridan, 36 but there was little sustained
discussion of improper supposition.

In logic texts, transumptio also appears in the sections devoted to the Topics.
In his De topicis differentiis Boethius, following a lost work of Themistius,
describes the topic ex transsumptione as the use of a thing which is better
known (such as justice in the republic) in place of something less well known
(such as justice in general), or of a more common word (such as “wise”) in
place of a word that is obscure or difficult (such as “philosopher”). 37 Boethius
was followed by William of Sherwood, 38 but some writers of logic texts,
including Peter of Spain, Lambert of Lagny (often called of Auxerre) and John
Buridan, replace the first case by the case of a word or phrase which is trans¬
ferred on account of a similitude, and here we find the standard examples of
“pratum ridet” and “litus aratur”. 39 John Buridan also added the cases of words
such as “lion” and “pig” predicated of human beings. In this context it is clear
that these logicians thought of transumptio in terms of metaphor.

Commentaries on the Sophistical Refutations

The fullest discussions of metaphor, most usually under the name transumptio ,
occur in textbook treatments of fallacies and in commentaries on the Sophisti¬
cal Refutations. In that work Aristotle first lists six types of fallacy which depend
on language, beginning with equivocation and amphiboly (I65b24-l66a6).
He then gives four examples of equivocation which were regarded as falling
into three groups. Following Boethius’s translation in AL VI 1-3 the first is
exemplified by the word “discere”, the second by the word “expediens”, and
the third by the words “sedens” and “laborans”. After some remarks about
amphiboly (I66a7-15), Aristotle then writes (l66al5-20): “There are three
modes of equivocation and amphiboly: one when either the phrase or the
name primarily signifies more than one thing, e.g. ‘piscis’ and ‘cams’; another
when we are accustomed to speak in that way; a third when words put

35) Walter Burley (1955), 2: 20-21. The full discussion is found 46-47. Vincent Ferrer (1977),
176-180.

36) Logica Cum sit nostra , in De Rijk (1962-1967), II.2, 447: 33-448: 8. Roger Bacon in De
Libera (1986), 287-289; Ockham (1974), 236-237; John Buridan (1998), 37-38.

37) Boethius (1978), 56-57.

38) William of Sherwood (1995), 124: 563-567.

39) Peter of Spain (1972), 74-75; Lambert (1971), 136; John Buridan (2001), 479-480.

322

E. J. Ashworth / Vivarium 45 (2007) 311-327

together signify more than one thing, but taken alone <signify> simply, e.g.
‘scit saeculum’.”

The first issue here is how to coordinate the two lists, given that each had
three members. In the literal commentaries of Albert the Great and Giles of
Rome the two are separated. In similar language, they report that the second
mode of equivocation is when a word signifies one thing prius (or principaliter)
and another thing ex consequenti , whereas the second mode common to equiv¬
ocation and amphiboly signifies one thing proprie and another thing per simili-
tudinem translationis or improprie for Albert and one thing principaliter by
virtue of imposition and another thing transumptive for Giles. 40 Giles added
that the second mode of equivocation was found among analogical terms (in
analogis). Later Ockham discusses the two lists separately, but treats the second
mode in much the same way. 41 Question commentaries and logic textbooks
tend not to make any distinction, and focus on Aristotles list of common modes
when they are discussing equivocation. Peter of Spain (99: 22-26) remarked
that one could take equivocation by itself or in relation to amphiboly, but that
there were just three modes in either case. Some early sources questioned in
what way the so-called common modes were in fact common, and suggested
that, whereas the first mode was genuinely common, the second mode related
to equivocation alone and the third to amphiboly alone, but this was never a
popular view. 42 On the other hand, it did raise the issue of how the three modes
were to be related to Aristotle’s discussion in Categories lal-5, where, according
to Boethius and other sources, there are just two main types of equivocal terms,
chance equivocals and deliberate equivocals. The standard answer was that
the first two modes dealt with signification, the subject of the Categories , and
were equivalent to chance equivocation and deliberate equivocation, whereas
the third mode dealt with modi significandi , consignification, or perhaps sup¬
position (depending on the source). 43 Thus, as Albert the Great explained,

40) Albert the Great (1890), 538b, 544a; Giles of Rome (1967b), f. 10 rb, f. 12 ra. Cf. Nicholas
of Paris in Ebbesen (1997), 170, dubium 2, and reply, 172-173.

41) Ockham (1979), 16-17 {proprie vs. improprie ), 20 {principaliter vs. ex translatione).

42) See Glose in Aristotilis Sophisticos Elenchos in De Rijk (1962-1967), I, 207 and Anonymus
Aurelianensis in Ebbesen (1979), 106-107. Anonymus Parisiensis in Ebbesen and Iwakuma
(1990, 79), says that the first two modes are common and the third is proper to amphiboly.

43) Albert the Great (1890), 538a {modi significandi)-, Giles of Rome (1967b), f. 10 rb {modi
significandi)-, Incerti Auctores (1977) Anonymus SF, 116-7 and Anonymus C, 309 {modi
significandi ); Peter of Spain (1972), 103 (consignification); Nicholas of Paris in Ebbesen (1997),
173, ad 5 (consignification); Ockham (1979), 24 (supposition) and Ockham (1974), 759-760
(supposition).

E. J. Ashworth / Vivarium 45 (2007) 311-327

323

equivocation is treated more widely in the Sophistical Refutations , which has to
treat of all the ways in which fallacy can arise, than in the Categories , 44

The tendency to identify the two lists of modes raises the important ques¬
tion of how the second mode of equivocation relates to transumptio , the sec¬
ond mode common to equivocation and amphiboly. The issue was complicated
by the introduction of analogy in the new sense whereby, because of attribu¬
tion, a term signifies one thing per prius and another per posterius or ex conse -
quenti. A number of sources pay little or no attention to this issue. Some,
particularly the authors of Question commentaries, refer only to analogy; 45
others refer, at least in effect, only to transumptio . 46 At least two English sources
belonging to the latter tradition refer in passing to analogical terms, 47 and in
his Summulae Roger Bacon included ens , which was a standard analogical
term, among his examples of transumptio. 48 At least two other English sources
solve the problem by placing analogy in the third mode. Thus William of
Sherwood (172) takes both “ridet” and “expedit” as examples of the second
mode in which words signify one thing principaliter and are transferred
{transumuntur) to another because of some similitude. The third mode (174),
illustrated by “laborans”, involves those dictiones whose “signification or
consignification is one intentio participated by many secundum prius et poste-
rius ” A similar solution is found in Simon of Faversham. 49 Other sources refer
to both analogy and transumptio without making any choice between them. 50

44) Albert the Great (1890), 538b-539a. See also Robert <Kilwardby> in Ebbesen (1997), 161.

45) See Anonymus Pragensis in Mure (1998), 89-91; Anonymus, Tres quaestiones de aequivoca-
tione, in Ebbesen (1998), 134; Anonymus G&C 611-11, Quaestiones in Sophisticos Elenchos in
Ebbesen (1998), 168-169, 176-177. Robertus Anglicus, Commentarium in Aristotelis Sophisticos
Elenchos in Ebbesen (1997), 193 rejects the view.

46) See various sources in De Rijk (1962-1967), I, 208 {proprie/improprie ), 290 {principaliter/ex
translation) , 499 {principaliter/translative et inproprie ); <Kilwardby>, Commentarium in Sophis¬
ticos Elenchos in Ebbesen (1997), 152 {principaliter/transumptive)\ Aiionymus e Musaeo in
Ebbesen (1997), 165 {simpliciter/ex transumptione)\ Eallaciaeadmodum Oxoniaeln Kopp (1985),
41-44.

47) Fallaciae breves in Kopp (1982), 264; Ockham (1979), 23: 27-30.

48) Roger Bacon in De Libera (1987), 242, §§448-449.

49) Simon of Faversham (1984), 85-86, 127-128. Curiously, on p. 80 he relates the per prius/per
posterius case to the second mode of equivocation, and refers to analogical terms.

50) See e.g. Dubia de aequivocatione in Ebbesen (1998), 127-128. For a more elaborate
classification, see Summe Metenses in De Rijk (1962-1967), II. 1, 475-476 and Dialectica Mona-
censis in De Rijk (1962-1967), II.2, 559: 20-28. The latter source is important because it empha¬
sizes that transumptio covers all kinds of trope: 561: 25-36. Cf. Ockham (1974), 758-759 for a
long list of types of translatio of which one type is metaphor (or transumptio : cf. 757: 39 ‘meta-
phorice et transumptive’).

324

E. J. Ashworth / Vivarium 45 (2007) 311-327

The Reduction of Analogy to transumptio , or the Reverse

More important for our purposes are those sources that refer to both analogy
and transumptio and explicitly reduce one to the other. One school, exemplified
by Peter of Spain and Lambert of Lagny, focussed on analogy, or at least on
terms that are said per prius et posterius , and reduced transumptio to analogy.
Peter remarked that one can view ridere as being said in a prior way of men
and a posterior way of fields. 51 In his Syncategoreumata he listed the ways in
which a word can signify different things, but kept secundum prius et posterius
and transumptive separate. 52 Unlike Peter of Spain, Lambert explicitly men¬
tioned analogical terms, which he assigned to the second mode of equivoca¬
tion. 53 The other school focussed on transumptio and saw analogy as secondary.
The Defallaciis wrongly attributed to Aquinas follows this school, and claims
that the multiplicity of analogical names which are said in a prior and a poste¬
rior way is reduced to the case in which a term signifies one thing principally
and another thing methaphorice siue transsumptiue.™

Few sources examine the relation between transumptio and analogy in
more depth. However, a certain Robertus and another Robertus, Robertus de
Aucumpno, both explain that even though analogy reduces to transumptio ,
there is a difference because in one case there is a true unity of nature and in
the other case there is only a resemblance, that is, a surface similarity. 55 A bet¬
ter explanation is found in the theologian James of Viterbo (d. 1308). He
argued that there is a crucial difference between analogy and metaphor. Where
there is attribution of one thing to another, there is a causal relationship of

51) Peter of Spain (1972), 100: 18-20 ‘Secunda species sive secundus modus equivocationis est
quando eadem dictio secundum prius et posterius significat diversa—101: 11-12 ‘Ad hanc
secundam speciem reducitur equivocatio ex transsumptione; 101: 20-23 ‘hec verba “ currit ” et
“ ridet” per prius significant ridere vel currere et per posterius florere vel labi , quia hec significant
ex propria impositione, ilia vero ex assuetudine.’

52) Peter of Spain (1992), 46.

53) Lambert (1971), 149 for analogy; 150 for the claim that fallacies arising from transumption
are reduced to the second mode.

54) Aquinas (1976), 406: 38-42, 53-55.

55) Robertus in Ebbesen (1997), 176: ‘Intellige etiam quod in analogo unum prius significatur
et alterum ex consequenti; et iste modus reducitur ad modum aequivocationis ex transumptione,
et ita non est differentia nisi [et] quod in analogo est unitas vere in significatis, in transumptione
non est vere sed solum similitudo.’ Robertus de Aucumpno in Ebbesen (1997), 187-188: ‘quia
in analogo unum primo significatur et alterum ex consequenti, reducitur ad modum aequivoca¬
tionis ex transumptione, et differt solum quia in analogo est unitas naturae maior in significatis,
in transumptione vero non unitas naturae sed similitudo.’

E. ]. Ashworth / Vivarium 45 (2007) 311-327

325

some sort, to one agent, or one end or one subject, but in translatio there is
only a relation of similitude. 56 This ontological issue, while not discussed at
any length by logicians, clearly lies behind some of the discussion about the
number of acts of imposition involved in deliberate equivocation. A causal
relation can be taken to justify a single act of imposition that covers two related
senses, though as we shall see it can equally be taken to justify the rejection of
such a single act.

Acts of Imposition

Ockham argued that just as chance equivocation involved two or more dis¬
tinct and unrelated acts of imposition, so deliberate equivocation of any sort
required two acts of imposition, so related that the second would not have
taken place had not the first already occurred. 57 More usually, logicians insisted
that deliberate equivocation required just one act of imposition, and this
allowed a principled distinction to be drawn between analogy and transump-
tio. For instance, Radulphus Brito writes in his Questions on the Sophistici
Elenchi that there are two sorts of analogy, the first when a term is imposed to
signify several things, one perprius and the other perposterius or ex consequenti,
and the second when a term is imposed to signify just one thing and is then
taken improperly (sumitur improprie) for another thing on the basis of some
similitude or other. 58 He uses homo pictus as an example, and says that this is
an improper kind of analogy.

The claim that transumptio is a subdivision of analogical terms, and perhaps
the only one that causes the fallacy of equivocation, is found in later authors,
notably in the fifteenth century logician and theologian Dominic of Flanders
(d. 1479), who enriches his discussion by an insistence on the role of concepts.
The term ridere comes to signify the concept flowering only by transference,
but a properly analogical term signifies either a concept containing some
ordering or an ordered group of concepts by virtue of its initial imposition. 59
Transumptio as a subdivision of analogy reappears in Gerard de Harderwyck
(d. 1503). In his commentary on the Categories he identifies deliberate equiv¬
ocation with analogy, and credits Albert the Great with the division of analogy

56) James of Viterbo (1983), 210.

57) Ockham (1974), 45. On two impositions, cf. Bacon (1988), 102; Bacon in Fredborg, Nielsen
and Pinborg (1978), 124.

58) Radulphus Brito in Ebbesen (1998), 217.

59) Dominic of Flanders (1967), sig. i 3 vb.

326

E. J. Ashworth / Vivarium 45 (2007) 311-327

into three: secundum proportionem ad unum (e.g. sanum ), secundum similitudi-
nem (e.g. homo pictus ), and secundum transsumptionem , (e.g. cursus said of a
man and of water). 60 Yet another approach is found in Cardinal Cajetan’s trea¬
tise De analogia nominum which was a supplement to his commentary on
the Categories. Cajetan echoes the early fifteenth century English Dominican
Thomas Claxton in distinguishing the analogy of attribution from the analogy
of proportionality, but where Claxton gave two undifferentiated examples of
the latter, visus and ridere f 1 Cajetan writes that the analogy of proportionality,
which is analogy in the Greek sense, has two modes, metaphorice et proprie ,
and ridere when said of a field or of fortune is his example of metaphor. 62
Cajetan thus restores the strong relationship between analogy in the Greek
sense, and metaphor.

All these authors accepted the existence of analogical terms, and those who
discussed imposition held that the single imposition of a term to signify per
prius etposterius was possible. For authors who rejected the possibility of such
a complicated initial imposition, but retained the view that only one act of
imposition was involved, appeal was made to transference from the proper
imposed signification, and analogical terms were either identified with meta¬
phors, or were excluded from the second mode altogether. The British school
is prominent here. In his Questions on the Sophistici Elenchi , Duns Scotus
argued that whatever relations of attribution hold in reality, it was impossible
to impose a single term to signify one thing per prius and another thing per
posterius , or to establish such a relationship between two distinct impositions
of a single term (336-7). Hence he argued that sanum was reducible to univo¬
cation, and that ens was reducible to the first mode of equivocation (344).
The second mode of equivocation did not include analogical terms but was
confined to translatio et transumptio a propria significatione ad impropriam, per
aliquam similitudinem , with ridere as the example (338-9). Four of his English
contemporaries, Thomas de Wyk, Walter Burley, Peter Bradlay, and the anon¬
ymous author of Quaestiones, retained the notion of analogy as applying to
the second mode of equivocation but made a clear distinction between the
analogy of being and the analogy of terms. 63 There is an analogical relationship

60) Gerard de Harderwyck (i486), sig. a iii vb-sig. a iiii ra.

61) Thomas Claxton (1943), 138-139.

62) Cajetan in Pinchard (1987), 120 §25.

63) Thomas de Wyk in Ebbesen (1998), 142; Burley, Super tractatum fallaciarum in Ebbesen
(2003c), 202; Peter Bradlay in Synan (1967), 282; Anonymus G&C 668, Quaestiones in Sophis-
ticos Elenchos in Ebbesen (1998), 289-290. The latter source allows that the word ‘ens’ is ana¬
logical, but only in a sense compatible with its being a univocal term.

E. J. Ashworth / Vivarium 45 (2007) 311-327

327

between the kind of being that is a substance and the kind of being that is an
accident, but, said Thomas de Wyk, this does not concern the act of signify¬
ing. The only analogical terms are those such as ridere which signify one thing
by imposition and another by transference. Burley added that this is the only
way in which a term can signify one thing per prius and another thing per
posterius (p. 201). Both Thomas de Wyk and Burley went on to explain that
the presence of just one act of imposition was crucial. 64 Imposition must be
totally adplacitum , but when there is transference, there is always a reason,
namely a likeness of relations, and so calling the lower part of a bridge its foot
cannot involve a second act of imposition.

Conclusion

Medieval logicians do not offer any sustained account of metaphor and how it
functions. Nonetheless, there is enough material to answer my initial ques¬
tions. First, while “transumptio” and “translatio” are quite often synonymous
with “metaphora”, one cannot take this as a general rule. Second, metaphor,
whether for the sake of ornamentation or out of necessity, was generally sub¬
sumed under deliberate equivocation. As a result, it was often identified with
analogy in the new non-Greek sense by logicians. This was obviously seen to
be compatible with the view that analogy in the Greek sense of a comparison
of relations was the basis for most metaphors. However, when “homo pictus”
was taken as an example of analogy and transumptio , the similarity involved
was one of external form rather than of relations. The most important contro¬
versies about the status of metaphor in relation to analogy were linked to the
issue of acts of imposition, and the question whether a single term could by a
single act of imposition come to signify secundum prius et posterius. Finally,
there are three features that serve to distinguish terms used metaphorically
from genuine analogical terms. Metaphors are not based on causal relations
among the objects signified; the metaphorical sense is not acquired by an act
of imposition; and when a word is used metaphorically it signifies something
other than what is properly signified by that word. When speaking of mead¬
ows, we say “laughing” and signify flowering.

64) Thomas de Wyk in Ebbesen (1998), 142-143; Burley in Ebbesen (2003c), 202.

BRILL

Vivarium 45 (2007) 328-342

VIVA

RIUM

www.brill.nl/viv

Scepticism, Demonstration and the
Infinite Regress Argument
(Nicholas of Autrecourt and John Buridan)*

Christophe Grellard

University of Paris 1 Pantheon-Sorbonne

Abstract

The aim of this paper is to examine the medieval posterity of the Aristotelian and Pyr-
rhonian treatments of the infinite regress argument. We show that there are some
possible Pyrrhonian elements in Autrecourt’s epistemology when he argues that the
truth of our principles is merely hypothetical. By contrast, Buridans criticisms of
Autrecourt rely heavily on Aristotelian material. Both exemplify a use of scepticism.

Keywords

scepticism, truth, evidentness, principles, infinite regress

The medieval reception of scepticism presents a most interesting example of
how a current or—to use the medieval term—a sect in ancient philosophy
was reconstructed. In some ways, the approach to sceptical questions in the
Middle Ages is a non-historical one, since it is hard to find what the ancient
philosophers called scepticism in this period. There were no sceptics in the
Middle Ages, in the sense of someone’s professing scepticism (except for
John of Salisbury). There was, however, a sceptical problem, understood as a
challenge to the theory of knowledge. This scepticism is a construction that
is not based on any historical foundation, and for this reason it should be
considered, above all, a breeding-ground of arguments (some of which are
ancient in origin, others genuinely medieval), a collection of challenges for
the theory of knowledge.

The main reason why medieval scepticism lacks a history and is reduced to
being a breeding-ground of arguments lies in the way ancient traditions of

0 Article translated by John Marenbon.

DOI: 10.1163/156853407X217803

C. Gre Hard / Vivarium 45 (2007) 528-542

329

scepticism were received in the Middle Ages. Without taking the traditions of
texts into account—establishing which texts were known and, most impor¬
tant, which were read, it is impossible to understand how a set of medieval
problems in a given area develops. With regard to scepticism, we need to
explain why this philosophical attitude was reduced to an extreme epistemo¬
logical thesis: ‘nothing can be known.

In this study, I link the problem of the sceptical tradition in the Middle
Ages to the question of a sceptical tradition in logic. In many ways, joining
together scepticism and logic may appear paradoxical, since in antiquity scep¬
tical discussion had among its aims to lay bare the dogmatic approach taken
in logic. None the less, it cannot be denied that in their project of suspending
assent the sceptics made use of logical techniques which may well have been
invented by others. To take account of sceptical arguments, Sextus Empiricus
uses an image that will reappear later in Wittgenstein—that of a ladder that
we push away once we have reached the top. There is, then, without doubt a
sceptical dimension to logic. In particular, sceptical interest in logic is linked
to the problem of the hierarchical organization of principles and the problem
of justification.

My aim here is to consider the case of a sceptical argument—that about an
infinite regress of justifications (‘Prove your proof’)—in order to unravel the
web of texts and to see how the argument was received and understood by
medieval thinkers. In particular, this argument enables us to reflect on the
status of the principles of knowledge. I shall look especially at Nicholas of
Autrecourt’s use of the infinite regress argument and will try to bring to light
some Pyrrhonian elements in his philosophy.

Ancient Traditions and Their Transmission to the Middle Ages

The transmission of the ancient sceptical tradition to the Middle Ages needs
delicate treatment, since medieval scholars seem to have known the main texts
of the tradition—Sextus Empiricus’s Outlines of Pyrrhonism and Ciceros Con¬
tra Academicos 1 —badly or slightly. Although there were many manuscripts of
the Contra Academicos in circulation, citations of it are very rare. John of Salis¬
bury (1125-1180), who vaunts his academic scepticism, seems to have known
only the Tusculanae\ and Henry of Ghent, although one of the few medieval
writers to know the Contra Academicos directly, contends that Cicero’s aim was

On this tradition, see Schmitt (1983) and Porro (1994).

330

C. Grellard / Vivarium 45 (2007) 328-342

to refute the Academics. 2 In the majority of cases, therefore, Cicero s text, and
the Academics’ positions more generally, were known in the Middle Ages
through the prism of Augustine. There is no need to stress the importance of
Augustine for philosophical reflection in the Middle Ages, and all his work
illustrates the importance of the sceptical problem for philosophy. As he recalls
in his Confessions , he had a personal experience of sceptical temptation, in the
form of anxiety about our ability to reach what is true, and it is this experience
which makes it an urgent necessity to refute scepticism. And yet Augustine
goes about doing so in such a way as to combine devising a refutation with
serving a useful purpose. Scepticism must be refuted in order to set up true
philosophy, but sceptical arguments can also have a propaedeutic value: they
can be used to refute false philosophies, such as empirical or materialist ones.

This Augustinian model sets the main lines for the medieval conception of
scepticism, but it is enriched by other, different traditions that are superimposed
on it. The first enrichment comes from the rediscovery from the twelfth century
onwards, through Aristotle as intermediary, of pre-Socratic traditions. 3 Two col¬
lections of texts are important. One the one hand, there is Book IV of the Meta¬
physics. The analysis of theses from Heraclitus, Protagoras and Democritus leads
to the examinations of a large number of arguments that can be considered
sceptical. Most often, reading these texts allowed an ontological foundation to
be given to sceptical doctrines. For medieval readers, it is because they judge, like
Heraclitus, that all things are movable or because, like Protagoras, they deny the
Principle of Contradiction, that the Academicians hold that nothing can be
known. On the other hand, the second group of texts is made up of the collec¬
tion formed by the beginning of the Posterior Analytics and Book II of the Meta¬
physics. These texts lead their readers to pose the question of what needs to be
demonstrated and, more generally, to be justified. It is in this context that medi¬
eval thinkers, in quaestio-commentai'ves, would once again examine the sceptical
problem: can we know anything? {Prior Analytics I) Can we grasp the truth?
{Metaphysics II).

For medieval thinkers, Posterior Analytics I, Chapter 3, is an important ele¬
ment in their reconstruction of scepticism, because here Aristotle explicitly
attributes the infinite regress argument to those who think that there is no
possibility of scientific knowledge (‘Certain people hold that there... does not
seem to be any scientific knowledge because we have to know the first princi-

2) See Schmitt (1972), 3-42.

3) On the trace of‘sceptical’ questions in Aristotle, see Long (1981).

C. Grellard / Vivarium 45 (2007) 328-342

331

pies’). 4 The infinite regress is directed against precisely those who judge that
there is no scientific knowledge apart from demonstration. As a result, this
argument sketches the portrait of a sceptic who demands a justification for
everything. But, at the same time, this argument is directly against the Aristo¬
telian definition of the syllogism as a demonstration that produces knowledge
(chapter 2; 71b 17: we know things through demonstration). In Posterior Ana¬
lytics I, 1-2 Aristotle requires that the premisses from which the conclusion is
deduced should be true (because there can be no knowledge of what is false),
anterior and better known. The transfer of justification which takes place in a
demonstration thus demands that the premisses have epistemic primacy. This
combination of demands (to know is to demonstrate, the premisses must be
better known than the conclusion) seems therefore to legitimate a sceptical
type of argumentation: there will necessarily be a regress of the premisses. The
impossibility of an infinite regress rests on the (psychological) fact that the
human mind cannot grasp an infinite chain of demonstration. It is also worth
noting that the infinite regress argument does not work on its own, but in
combination with the hypothetical mode. If we stop in an arbitrary way at a
principle, this principle will not be known and so will not take part in the
transfer of justification (of evidence and truth) that is expected of a syllogism.
Either the foundations of science are undiscoverable, or else we must admit
that they are merely hypothetical: either way we have to give up an absolute
definition of knowledge.

Aristotle gives a two-fold reply to this sceptical argument, a logical one and
an epistemological one. The logical answer is found in Book I, 19-23, and it
aims to prove that the requirements of essential predication make a regression
impossible whether it is based on the middle terms or the first and last
(extreme) ones. The logical solution (it might be called the negative one) con¬
sists in rejecting the possibility of infinite regress by showing that the extreme
terms reciprocally limit each other: given the meaning of each of the terms,
there is only a finite number of predications possible. Thus Chapter 22 shows,
in a first argument, that infinity is excluded both in essential and non-essential
predication. If infinite regress were allowed where essential predication is
concerned, it would be impossible to construct a definition. But we are in fact
able to give (essential) definitions. This argument is therefore constructed as a
reductio (82b37-83al). Then Aristotle generalizes his point to cover all predi¬
cations, that is to say, non-essential predication too.

4) Aristoteles latinus (Translatio Iacobi), Analyticaposteriora , I, 3, 72 b 3—72 b 23, in AL IV.

332

C. Grellard / Vivarium 45 (2007) 328-342

The second argument is linked more directly to the problem raised in
Chapter Three (83b35-84a5). The finite nature of the predication of middle
terms is presented as a necessary condition for it to be possible to make a dem¬
onstration and so to gain scientific knowledge. Aristotle reasons here as if the
problem raised by the sceptics (and which resists both on infinite regress and
the hypothetical mode) had been resolved. And so Aristotle seems ultimately
to presuppose that any solution to the problem of infinite regress is found on
the epistemological side. This is effectively the conclusion of this group of
chapters (84a29-b2). It has just been proved that a demonstration is made by
placing a middle term between two extreme ones, and that this placing cannot
go on and on to infinity. We know, therefore, negatively that not every propo¬
sition is demonstrable because not every proposition has a middle term. But it
remains to explain, positively, how the principles can be known. How can one
know truths that are indemonstrable because they lack a middle term? What
logic allows us to affirm is that not every thing can be demonstrated. This is
the first step in refuting the adversaries mentioned in Chapter 3, but it remains
to explain how these indemonstrables can be known, if one wants to escape
from the hypothetical mode. Aristotle is therefore led to defend the idea that
certain truths can be established without arguments. Ultimately, then, found¬
ing a system supposes finding foundations outside the system. Thus the con¬
clusion of Chapter 22 echoes that of Chapter 3.

Aristotle s solution is known, but it is debated because of its ambiguity. As
early as Chapter 3, Aristotle introduces the idea of a ‘principle of science that
makes us capable of knowing definitions/ For a medieval reader, the sense is
different, and in a certain way more precise and reductive: principium scientie
esse quoddam dicimus , in quantum terminos cognoscimus. Following this transla¬
tion, the principles which put a stop to infinite regress are those principles that
are known from the examination of their terms—that is to say, analytic prop¬
ositions. This is not, however, the essential point of the reply. This reply is
found in Posterior Analytics II, 19, which seems to be a sort of appendix to the
main body of the treatise. Here, after ruling out innate knowledge of princi¬
ples, Aristotle seems to hesitate between, on the one hand, an empirical pro¬
duction of principles, through induction, which would start from sensation
and rise up to the universal thanks to memory and experience; and, on the
other hand, an intellectual intuition—a direct grasp of principles. In this
second alternative, what is involved is the intellectual faculty of apprehen¬
sion of simples, which is completely free from error ( On the Soul III,6;
430a26—430b30). For medieval readers, then, there were at least four avail¬
able alternative solutions to the problem of infinite regress: (1) knowledge

C. Grellard / Vivarium 45 (2007) 328-342

333

through the analysis of terms; (2) knowledge through induction; (3) innate
knowledge (Plato’s position, in the form in which Aristotle criticizes it and
it is taken up again by theories of illumination); (4) intuitive knowledge
(intellectus).

There is a final tradition that should be considered, although its impact
remains difficult to evaluate—that of the Outlines of Pyrrhonism. It is well
known today that manuscripts of a Latin translation of this text circulated at
the end of the thirteenth century or beginning of the fourteenth century.
Three manuscripts survive but we know nothing about their owners. It is
therefore difficult to evaluate the influence of this text in medieval epistemol¬
ogy, but it cannot be completely excluded. 3

We know that in Outlines of Pyrrhonism, Sextus Empiricus recounts ten modes
or tropes which were attributed to the ancient sceptics (Aenesidemus, in fact)
and five modes to the new sceptics (Agrippa, according to Diogenes Laertius).
They are procedures designed to lead to a suspension of judgement by show¬
ing the undecidable character of things in the absence of any criterion. What
should be noted is that, whereas the ten tropes are based above all on relativity
(of the objects of knowledge and of the subject), the five tropes of Agrippa are
based, rather, on logical procedures developed throughout the course of antiq¬
uity and developed here in a systematic way.

Agrippa’s aim is to show that every dogmatic thesis, i.e. every attempt to
find a criterion for truth, is necessarily guilty of an error: contradiction (a
disagreement between two theses, p and not-/>), infinite regress, relativity,
being hypothetical, circular reasoning. As Jonathan Barnes has shown, these
five modes function together as a network and form what he calls a web. 5 6
Take a thesis p. The first mode allows us to put forward that there have been
other philosophers who have held -/>, especially because of the relativity of
knowers and the circumstances in which they know (third mode). And so a
proof q is necessary, such that q -> p , but this proof itself presupposes another
proof: either this proof is of the same nature as the proof of p and we enter a
vicious circle (fifth mode), or it is merely hypothetical (4th mode). Through
its paradoxical alliance with scepticism, the critical force of logic becomes
clearly apparent, and this was something medieval thinkers knew very well.

5) Recent work on the availability of Sextus Empiricus in the Middle Ages includes Cavini
(1977), Mutschmann (1891) and Wittwer (2002).

6) Barnes (1990), 113-149.

334

C. Grellard / Vivarium 45 (2007) 328-342

Looking more particularly at the second mode, its way of functioning is
easy to grasp. As Sextus Empiricus stresses, it comes into play when it is neces¬
sary to ‘bring about conviction (pistis, fides)' (I, 164-6), that is to say, to justify
the truth of a thesis. The repercussions of this need for justification bring
about the constitution of an infinite chain of justifications. But such a chain is
impossible, and it must be admitted that the initial thesis is unjustifiable. Sus¬
pension of assent should follow. But one can ask why it is impossible to admit
an infinite chain of justifications. Sextus’s reply is very clear and revealing. We
would have no point of departure, no principle or foundation (‘having noth¬
ing from which we can begin to set up’: non habentibus nobis unde incipiamus
constructionem , where constructio translates archometha). In speaking of‘con¬
struction’ the Latin is more explicit: what are envisaged are foundationalist
theories of knowledge, which demand that, in order to be justified, all knowl¬
edge must have a foundation, that is to say, must draw its evidence or its truth,
by transfer, from a basic belief or a belief which has a special status—the foun¬
dations that sustain the whole structure. What Sextus stresses is that, from the
logical point of view, there is no basis for supposing that there are any beliefs
of this sort.

Two observations are necessary. First, Sextus’s argument seems to a great
extent to rely on confusing the logical and psychological levels. Sextus thinks
of demonstrations as psychological events that occupy a finite period of time.
It is this genetic way of seeing it that rules out infinite regress (II, 85). The
temporal nature of proof is such that it leads at one time or another to a
vicious circle or to an arbitrary foundation. Second, however, it is undeniable
that Sextus poses clearly the problem of the status of principles in a demon¬
stration. How can it be legitimate that the logical point of departure for a
demonstration is outside the criteria of truth demanded for a demonstration?
What were the medieval repercussions of this challenge?

The Status of the Principles and the Use of Infinite Regress in Nicholas
of Autrecourt

The infinite regress argument encourages in particular reflection on the status
of the principles of knowledge: that is to say, to determine what a principal is
and to reply to one of the questions from which scepticism is made up: what
can be demonstrated? As shown above, this question occurs several times in
Aristotle, in the Posterior Analytics and, especially in the Metaphysics II and IV,
with regard to the justification of the first principle, the principle of non-

C Grellard / Vivarium 45 (2007) 328-342

335

contradiction. The idea that scepticism is linked to an ignorance of logic or to
sophistical practices is fairly common in the Middle Ages. 7 Nicholas of Autre-
court’s contribution to this controversy is interesting, because he uses an argu¬
ment close to Sextus’s to demonstrate that the truth is no more than a
hypothesis which is guaranteed only by indemonstrable principles.

Nicholas’s theory of knowledge is foundationalist par excellence : it is a the¬
ory that supposes that there are basic self-justifying beliefs, which serve as the
principles or foundations for the whole structure of knowledge that can be
inferred from them. The basic items of knowledge are of two sorts: immediate
perceptions—as the result of a principle according to which all that appears
is true (a principle which is described as ‘probable’); and the complex
principles—analytic propositions that can be reduced to the first principle or
principle of non-contradiction. 8 There are, then, two meta-principles under
which can be grouped the whole collection of items of basic knowledge, from
which other knowledge can be inferred. It is characteristic of these principles
that they cannot be false. I shall leave aside here the question of whether the
empirical principles (veridical perception) are evident, and confine myself to
the logical principles. As a result of his foundationalism, Nicholas is immedi¬
ately concerned with the question of regress when providing justification.
He never at any time treats the intellect as a faculty of principles of knowl¬
edge. There are therefore two problems: what are the logical principles of
knowledge—in what sense are they conditions for the possibility of truth?
How are they known—how do we know that they do indeed serve as a foun¬
dation for knowledge?

In this context, Nicholas uses the infinite regress argument, but he never
tries to refute it. The regress is accepted as a challenge to any foundationalist
theory of knowledge, because if it is effective, it will make all knowledge
impossible. Regress to infinity is thus used to prove the necessity for a founda¬
tion. Nicholas’s thesis—a more radical version of Aquinas’s—is that all our
knowledge must be founded on the first principle, which is the principle of non¬
contradiction. This thesis is Aristotelian in origin, because in Metaphysics IV,4
(1006a5-10) Aristotle defends the idea that the principle of non-contradiction
is the most certain of all. As a result, for Nicholas all certainty is that conferred
by this first principle to which all other principles must be reduced, along
with the conclusions of our demonstrations, where ‘reduce’ means ‘show the

7) Thomas Aquinas, In Met. IV, 1. 6, n. 12.

8) O’Donnell (1939) (Exigit ordo = EO) 235: 6-9; cf. Grellard (2005), chapters 2 and 3.

336

C. Grellard / Vivarium 45 (2007) 328-342

foundation of’. In this sense, the first principle functions as a sort of meta¬
principle with regard to analytic propositions, which can be taken in each case
as the principle for a particular demonstration. It is here that Nicholas takes
up the infinite regress argument in order to show that every principle of a
demonstration is not sufficiently justified but still needs to be reduced to the
first principle. Section 8 of the second letter to Bernard is thus very clear: the
conclusion of a syllogism must be reduced to the first principle either imme¬
diately, or mediately, that is to say, in such a way that the principle of this
conclusion is the conclusion of another syllogism. But this reduction has to
stop somewhere or it will enter an infinite regress. 9 Every proof contains the
first principle virtually, and a regress of proofs will simply make this ultimate
foundation actual. Nicholas seems therefore to presuppose here that only the
first principle truly answers to the need for indemonstrability and it alone can
function truly as a principle.

What is the meaning of this primacy of the principle of non-contradiction
and how is it known? Nicholas does not really explain how it is known. He
seems to be content with the fact that it is indemonstrable, because to demon¬
strate it necessarily leads to begging the question. Nicholas admits the princi¬
ple of non-contradiction simply because it is a necessary condition for debate.
Nicholas insists strongly on the foundational character of the first principle. It
constitutes, first of all, the condition for the possibility of all discourse, and in
this sense, it is presupposed by all philosophical dispute. This primacy of the
principle of non-contradiction is seen in two specific ways. First, it enjoys a
primacy that Nicholas calls negative. Nothing comes before this principle,
in the sense that it cannot be the conclusion of a syllogism—it cannot be
demonstrated—nor reduced to another principle. On this point, Nicholas is
in agreement with Aristotle, for whom the principle of non-contradiction can¬
not be demonstrated but is required necessarily by every demonstration. 10 The
primacy of the principle of non-contradiction is also described in a positive
way. Every conclusion of a demonstration must be able to be reduced to the
evidentness of the first principle, which is what puts a stop to an infinite
regress. This idea of reduction is important because it underlines the role of
the first principle in a theory of knowledge: that of providing epistemic
justification. It is indeed in the very context of the justification of our complex

9) Nicholas of Autrecourt, Second letter to Bernard, § 8. For a fuller account of my views, see
my Introduction (esp. 29-41) to Nicholas of Autrecourt (2001).

10) Cf. Aristotle, Metaphysics , IV, 4, 1006 a 1—1006 a 25 (Aristoteles latinus , XXV, 2, 66-67).

C. Grellard / Vivarium 45 (2007) 328-342

337

knowledge that Nicholas introduces this idea of reduction to the first principle.
Certainly, Nicholas accepts (§2) that any analytic proposition, since it is
known by a simple analysis of the terms which reveals that the predicate is
included in the subject, is evident and can function as the principle of a dem¬
onstration (EO 236:40-2). But what reveals the inclusion of the predicate in
the subject is the fact that the predication is not contradictory. And so, in the
final recourse, it is the first principle that once more appears as the ultimate
foundation.

One problem remains, however, in suspense. Who authorizes us to say
that our principles are true and that there exists something true independ¬
ently of our demonstrations, given that the first principle guarantees only
the evidentness of the conclusions of demonstrations? In fact, for Nicholas,
truth remains precisely an indemonstrable hypothesis (and in a certain way
an unnecessary one), and it is the infinite regress argument that allows him
to establish this.

His point of departure is as follows (EO 233:39-49). If the intellect can say
about a thing that it is true, and so by virtue of the first principle, the opposite
of this thing is not known. Conversely, what is evident, and so known, must
be admitted as true. Truth is therefore a simple hypothesis of which, at best,
an indirect proof can be given, by reductio. Nicholas reasons as follows:

1. \{p is true (= if it is posited as an hypothesis: it is a fact that p / it is thus),
it cannot be known that -p. [to know = df clear and evident apprehension;
by application of the principle of non-contradiction]

2. If it is known that />, then p is true, [converse of (1)]

3. If (If I know that />, then is true), then nothing is certain, that is to say,
the principle of non-contradiction is not valid (there is no knowledge),
[proof by reductio : negation of (2), that is of the thesis that one wants to
demonstrate]

4. Something is certain, [proof by the principles: the principles are appre¬
hended in a clear and evident way, knowledge of the nominal definition of
the terms is enough to gain the intellect’s assent] (See EO 235:6-9)

5. Therefore, it is not the case that (If I know that p , then - p ). [application of
modus tollens to (3): p -> q\ ^q\ therefore ^ p]

6. I know that p , etc. (2) [proof by reduction negation of the thesis, negation
of the negation, and so thesis (2)]

It seems clear then that Nicholas does not have a direct proof of the truth of
knowledge founded on the principles. It merely has to be admitted that there

338

C. Grellard / Vivarium 45 (2007) 328-342

is a link between evidentness and truth, that is to say, we must put forward the
hypothesis that what is evident is true.

Indeed, Nicholas proves just afterwards that it cannot be established directly
that our evident principles are true (EO 237:1-7). We must admit as a princi¬
ple that there is truth. For if we want to prove that a proposition p is true, we
are faced with the following dilemma, the premisses q and r which allow us to
prove that p is true need to be either (1) true or (2) evident. If (1), then we still
have to show that q and r are true, otherwise we are begging the question (or
else we are using the hypothetical mode). But, in order to prove the truth of
these premisses, they have to be inferred from other true premisses, ^’and r\
and so on. But then we shall have an infinite regress. And so we must say
(2) that p is true because the premisses which demonstrate it are evident, that
is in themselves. But we are guilty of begging the question by granting what
we wish to demonstrate, that is to say, the link between truth and evidentness.
In this way the conjunction of the three sceptical modes—infinite regress,
hypothetical reasoning and the vicious circle—allow Nicholas of Autrecourt
to show that truth is an indemonstrable hypothesis.

Buridan’s Criticism of Nicholas of Autrecourt: The Multiplicity
of Principles

Nicholas’s contemporaries, as is well known, were far from accepting the way
in which he radicalized the need for foundation by the first principle. In par¬
ticular, his colleague in the Faculty of Arts, John Buridan, criticized the idea of
a unique first principle and proposed instead the idea of a multiplicity of prin¬
ciples of demonstration. 11 What is interesting is that, in refuting Nicholas,
Buridan is faced by the same questions about how the principles of knowledge
can be known, and he too has to face the infinite regress argument.

Buridan’s Summulae dialecticae VIII,5 opens, in a rather annunciatory and
programmatic way, on the very question of infinite regress. Borrowing almost
literally Aristotle’s argument in Posterior Analytics 1,3, he deduces from it the
need to prevent a demonstration from stopping at an indemonstrable premiss.
One or several principles are therefore necessary. 12 Nicholas of Autrecourt’s
position is, moreover, recognized as central with regard to this question about

10 See especially King (1987), Zupko (1993) and Grellard (2003), chapters 8 and 9.
12) John Buridan (1999) {Summulae dialecticae = SL), VIII, 5, 1.

C. Grellard / Vivarium 45 (2007) 328-342

339

principles, and it is presented as one of the solutions to the regress argument.
Buridan wants to show that it is impossible that there is only one principle.
The argument is primarily logical and it rests on the syllogistic form of all
demonstrations. Because the conclusion of a syllogism is demonstrated by two
premisses, either these two premisses are indemonstrable, or both or at least
one of them must itself be demonstrated. But, given the impossibility of an
infinite regress, it must be admitted that we shall arrive at a syllogism com¬
posed of two indemonstrable premisses (SL VIII,3,2). As a result, since these
indemonstrable premisses are principles, according to Aristotle himself, it
must be admitted that there is not just one principle. It is clear that here
Buridan has in view the argument based on the syllogistic form put forward
by Nicholas in Section 8 of his second letter to Bernard. 13 It is therefore inter¬
esting to realize that Buridan uses the infinite regress argument, just like
Nicholas, to defend the necessity of a foundation. But a distinction between
the two philosophers appears immediately, because for Buridan the principles
are many. Here Buridan is placing the first block in his defence of both a par¬
tial foundation of knowledge and a multiplicity of principles. Buridan must
still, however, explain why Aristotle says of the principle of non-contradiction
that it is the most certain and implies that its primacy is unique. The answer is
two-fold. From the strictly logical point of view, Buridan distinguishes being
primary in respect of commonness, and being primary with regard to the evi¬
dentness of primacy, in the sense of indemonstrability. The principle of non¬
contradiction, being the first principle, is the most common principle: it can
be used in a demonstration of any type whatsoever (whereas the sciences have
to use their own principles). Similarly, it is the most evident principle. But,
according to Buridan, what is brought up by the infinite regress is primacy in
the sense of indemonstrability (first being understood as immediate, the
absence of any anterior premiss). We have here, therefore, a sharp distinction
between justification and demonstration. For Nicholas, the most certain is the
best justified and so what makes any demonstration legitimate, whereas Buri¬
dan defends the legitimacy of demonstration that does not require the highest
degree of certitude. Buridan therefore accepts the negative primacy of the first
principle in the way that Nicholas understands it (the first principle is not the
conclusion of a demonstration), but not the positive primacy (the first princi¬
ple is the premiss of every demonstration). 14

13) See above, p. 336.

14) See Nicholas of Autrecourt, Second letter to Bernard §2, and cf. n. 10.

340

C. Grellard / Vivarium 45 (2007) 328-342

As well as this exegetical answer, Buridan gives a logical and epistemological
one. The idea that knowledge is one because it is founded on a unique princi¬
ple should be rejected, because there as many principles and middle terms as
there are conclusions. There is an infinity of conclusions and so there is an
infinity of principles. Buridan proves the infinity of conclusions by referring
to mathematical examples where the properties of an infinity of figures can be
demonstrated, and from there he draws the conclusion that the principles are
infinite, since for every object of demonstration there must be infinitely many
different subjects, not subordinated according to predication, and so there is
an infinity of indemonstrable premisses (SL VIII,5,2). The question that can
be posed, however, is to know whether in this case Buridan does not deprive
himself of the logical solution to the infinite regress which Aristotle proposed
and which is based on the reciprocal limitation of the extreme terms. If there
is an infinity of middle terms, there will be an infinity of subjects and predi¬
cates. Will there not be an infinite regress of middle terms, so that it will
become impossible to reach the principles. In other words, does not the
defence of the infinity of principles and middle terms lead to the reintroduc¬
tion of the infinite regress? Buridan’s reply is extremely interesting. There is an
infinity of subjects and predicates, but for a given subject and predicate, there
is not an infinity of possible predications. Buridan brings out the full meaning
of the idea of the reciprocal limitation of the extreme terms. The whole collec¬
tion of terms is infinite, but the construction of a predication is not a matter
of choosing among infinite combinations. The subject of a science (the genus,
in Aristotelian terms) necessarily limits the principles. The theoretical possibil¬
ity of an infinite regress thus is seen to be purely formal, and in practice dem¬
onstrations always makes use of principles constructed on the basis of middle
terms determined by the subject of the science.

It still remains, however, for Buridan to explain how the principles are
known. 15 For this purpose, Buridan puts forward an empirical reading of the
Aristotelian solution. First, analytical and empirical principles must be distin¬
guished. 16 Analytical principles can serve as a model, given that it is impossible
to refuse to assent to them. The intellects natural inclination to approve them

15) See De Rijk (1994) and Zupko (2003), esp. chapter 12.

16) John Buridan, Quaestiones in libros analyticorum posteriorum , II, q. 11, unpublished tran¬
scription by H. Hubien: ‘Deinde etiam notandum est quod duplicia sunt principia indemonstra-
bilia. Quaedam sunt quorum rationes terminorum se manifeste includunt aut manifeste
excludunt scito quid nominis. Alia sunt principia quorum rationes terminorum nec se manifeste
includunt nec manifeste excludunt.’

C. Grellard / Vivarium 45 (2007) 328-342

341

is therefore the sign of our more general inclination to approve what is true.
Conversely, the second type of principal is not immediate (that is to say, assent
is not chronologically immediate, but the proposition is immediate in that
it is indemonstrable), but presupposes an inductive process, starting from
something that is sensibly given, according to the genetic scheme set out by
Aristotle in the Posterior Analytics 11.19. Buridan emphasizes the naturalist
aspect of his epistemology here by indicating that it is as natural for the intel¬
lect to approve the principles (analytic or empirical) as for fire to burn. It is
therefore clear that Buridan held that it was necessary for our knowledge to be
founded on principles. In multiplying these principles, however, he makes this
foundation partial and determined by context. It seems to him necessary to
reject Nicholas of Autrecourt s reductionist ideal. What in fact emerges is that,
for Buridan, the question of the beginning, the point of departure, of knowl¬
edge is not simply a question of logic, but it is a question of fact, which
depends on each particular science. We have an example of this relativity of
foundations in the particular case of the circular syllogism. A syllogism is never
absolutely circular—this type of circularity is rejected by Aristotle at the very
point after he has rejected infinite regress. And yet circularity relative to a
previous syllogism is possible, where from the conclusion and the converse of
one premiss, the other premiss is inferred. This particular case underlines the
importance in the enterprise of justification of the coherence of our beliefs
amongst themselves, as opposed to an exclusively foundationalist conception
which is open to infinite regress. Buridan therefore insists especially on our
capacity to construct proofs each time it is necessary, based on principles
adapted to the object of study.

Conclusion

It can therefore be seen clearly how the treatment of infinite regression by Nich¬
olas of Autrecourt and John Buridan contributes to clarify their different con¬
ceptions of epistemic justification and of the role and status of the principles of
knowledge. Buridan clearly depends on the Aristotelian treatment of the ques¬
tion of infinite regress in his approach, and he gives it in the main a logical
answer. By contrast, whether he took his information from Aristotle or from
Sextus, 17 it is interesting to note that Nicholas uses the infinite regress

17) See Sextus Empiricus, Informationespirronarum , II, 85, Paris BN lat 14700, f. 104 vb: ‘Dis-
sonancia igitur existentia a dogmaticis de vero, quia quidam dicunt utique esse aliquid verum,
quidam autem dicunt nichil esse verum, non contingit dissonancia iudicare (...); veram autem

342

C. Grellard / Vivarium 45 (2007) 328-342

argument to establish the necessity of a first principle (the foundation of all
knowledge) and to defend the primacy of what is evident over what is true,
which is reduced to a mere hypothesis. We cannot, however, speak of Nicho¬
las’s ‘Pyrrhonism’, because a large part of Pyrrhonian procedure (the idea that
epoche leads to ataraxia ), is completely unknown to him. Even the idea of
Nicholas’s ‘scepticism’ is highly arguable. 18 None the less, it remains true that
some of his positions produce an effect of scepticism, and that some Pyrrho¬
nian elements can be identified in his work. But what I have wanted to empha¬
size most of all is this: given that there was a sceptical tradition in the Middle
Ages, which was partially inherited from Antiquity and remodelled on the
basis of characteristically medieval problems and tools, its principal aspect is
methodological. Thus, taking this particular sceptical argument as our point
of departure, we see that the sceptical problem in the Middle Ages, rather than
being an epistemological doctrine that holds that nothing can be known, is an
epistemological stimulus that leads thinkers to face the positions they defend
with theoretical difficulties that require them to clarify their thinking. And so
in the Middle Ages scepticism is more than a thesis or a systematic philosophy:
it is a breeding-ground for arguments which make philosophers think.

demonstrationem esse dicens meam quod per se invicem est incidit rationem, et demonstratio-
nem petetur veram oportet ipsam esse et illius aliam et usque in infinitum. Impossibile est autem
infinita demonstrare , impossibile est ergo scire et quod sit aliquid verum.’

18) I return to this question in Grellard (forthcoming).

BRILL

Vivarium 45 (2007) 343-359

VIVA

RIUM

www.brill.nl/viv

Theory of Supposition vs. Theory of
Fallacies in Ockham

Catarina Dutilh Novaes

Leiden, The Netherlands

Abstract

I propose to examine the issue of whether the ancient tradition in logic continued to be
developed in the later medieval period from the vantage point of the relations between
two specific groups of theories, namely the medieval theories of supposition and the
(originally) ancient theories of fallacies. More specifically, I examine whether supposition
theories absorbed and replaced theories of fallacies, or whether the latter continued to
exist, with respect to one particular author, William of Ockham. I compare different
parts of Ockhams Summa Logicae , namely II1-4 (on fallacies), and the final chapters of
part I and first chapters of part II (on supposition). I conclude that there is overlap of
conceptual apparatus and of goals (concerning propositions that must be distinguished)
in Ockhams theories of supposition and of fallacies, but that the respective conceptual
apparatuses also present substantial dissimilarities. Hence, theories of supposition are
better seen as an addition to the general logical framework that medieval authors had
inherited from ancient times, rather than the replacement of an ancient tradition by a
medieval one. Indeed, supposition theories and fallacy theories had different tasks to
fulfil, and in this sense both had their place in fourteenth century logic.

Keywords

Ockham, supposition theory, theory of fallacies, distinguishing propositions

When reflecting on the traditions of ancient logic in the Middle Ages, one
may even go as far as asking: is there such a thing as ancient logic in the Mid¬
dle Ages? The general problem can be put as follows: true enough, medieval
logical theories were almost always in some way or another inspired by ancient
logical theories. But two hypotheses suggest themselves:

- Medieval theories replaced, or at least merged with, the ancient theories,
and thus the latter ceased to exist as such;

DOI: 10.1163/156853407X217812

344

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

- Ancient theories continued to exist alongside with the medieval new
theories.

In fact, historical evidence indicates that it is the second hypothesis that best
describes the actual developments of that period; 1 that is, ancient logic contin¬
ued to be developed as such, and was not entirely replaced by the medieval
novelties. But the interaction between “new (medieval) logic” and “old (ancient)
logic” 2 deserves further analysis.

I propose to examine this issue from the vantage point of the relations
between two specific groups of theories, namely the medieval theories of sup¬
position and the (originally) ancient theories of fallacies. Of course, we are all
familiar with De Rijk’s thesis: theories of supposition were developed in the
twelfth and thirteenth century within the framework of fallacy theories. 3 But
the issue as to whether supposition theories absorbed and replaced theories of
fallacies, or whether the latter continued to exist, is worth examining in more
detail. More specifically, I intend to analyse the state of affairs in the four¬
teenth century with respect to one particular author, William of Ockham. I
compare different parts of Ockham’s Summa Logicae , namely III-4 (where he
presents his analysis of the theory of fallacies), and the final chapters of part I
and first chapters of part II (where the notion of supposition plays a central
role); to a lesser extent, I also refer to Ockham’s commentary on the Sophistici
Elenchi (vol. Ill of Ockham’s Opera Philosophica). A

Preliminary Considerations

My aim is to compare Ockham’s theories of supposition and of fallacies. How
can one compare two theories? Clear criteria of comparison are needed. For

'> Cf. Ebbesen (1982).

2) Notice that this distinction does not refer to the “ logica vetus / logica modemorum distinction.

3) Cf. De Rijk (1962, 1967). For the present purposes, it is irrelevant whether De Rijk’s thesis is
correct or not; its role in the discussion presented here is merely that of motivating a closer
inspection of the relations between theories of supposition and theories of fallacies.

4) There seem to be methodological reasons for focusing on Summa III-4 rather than on Exposi-
tio Elenchorum : a Summa is obviously more likely than a commentary to present a systematic
account of the theories and doctrines defended by a given author; the Summa in its totality
contains both Ockhams theories of supposition and of fallacies, allowing for textual unity in the
comparison; and finally, in a commentary such as the Expositio , authors tend to be more “obedi¬
ent” towards the text being commented on, instead of presenting their own theories. This being
said, in the Summa one does not find meta-remarks on the nature of the theories being discussed,

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

345

the present investigation, I adopt two main criteria, which appear to be two
crucial aspects of any theory:

- The goals of the theory, the results it is expected to accomplish;

- The (conceptual and technical) apparatus used in the theory.

Two theories may have the same goals, while the conceptual apparatus used to
attain these goals is very different; moreover, two theories may use a very similar
apparatus, but to perform different tasks. In what follows, I employ the notions
of goal and apparatus in order to compare the two theories in question.

Concerning Ockham’s theories of supposition and of fallacies, one cannot
help noticing the trivial fact that each of them is treated in different parts of
the Summa (the same holds for Buridan’ Summulae de Dialectical by the way).
So it might seem that I am just stating the obvious—that Ockham’s theory of
supposition and his theory of fallacies are indeed different theories. But given
that a strong case has been made for the historical hypothesis that theories of
supposition have emerged from theories of fallacies, the issue of the bounda¬
ries between them in an author such as Ockham is worth investigating. In
other words, that these are different theories but with points of contact may
be a rather uninformative claim, but showing where they differ and which are
(some of) these points of contact in the case of Ockham seems to be an
endeavour worth being undertaking.

Similarities

Assuming that De Rijk’s thesis is at least to some extent historically correct,
and that theories of supposition did develop within the framework of theories
of fallacies, one is inclined to expect significant similarities between them.

The issue of conceptual priority of one theory versus another is also at
stake: it is often the case that medieval philosophers would make use of well-
established Aristotelian concepts to justify and provide foundations for their
own theories. So one could expect the theory of fallacies to serve as foundation
for the theory of supposition, but the situation is often more complex than this.

while such remarks do help clarifying them; therefore, I also turn to the Expositio in order to
include some of those meta-remarks in the discussion.

5) In Buridan’s Summulae , supposition is the subject matter of the 4th treatise and fallacies of
the 7th treatise.

346

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

Equivocation—Modes of Supposition, Temporal Propositions

The most conspicuous point of contact between Ockhams theory of fallacies
and his theory of supposition concerns the fallacy of equivocation, in particular
the distinction between personal, simple and material supposition and the
third mode of equivocation. The third mode of equivocation is the contextual
mode of equivocation, and concerns terms that, taken individually, are not
equivocal, but which, when put in a given propositional context, can stand for
different things. According to Ockham, certain propositions allow their sub¬
ject term to have more than one kind of supposition—and therefore must be
distinguished—, and that corresponds neatly to the third mode of equivoca¬
tion. In fact, this mode of equivocation serves as conceptual justification for
the distinction between the three kinds of supposition, as the passage below
shows (notice the term 'penes):

But in the proposition “Man is a species”, because “species” signifies the intention of the
soul, it can have simple supposition. And it is a proposition to be distinguished according
to [penes] the third mode of equivocation, in that the subject can have personal or simple
supposition. (I, 65, 13-17)

Interestingly, the general rules determining when a term can have a supposi¬
tion other than personal supposition in a proposition, which had been formu¬
lated in I, 65, are stated again in the chapter dedicated to the third mode of
equivocation:—

There is a rule that, when of the major and minor terms of a proposition one is a name of
first intention, without a universal or particular sign, and the other is a name of second
intention, that proposition should be distinguished according to the third mode of equivo¬
cation, in that the name of first intention can supposit simply or personally, and as a result
of this the fallacy of equivocation can occur in the argument (III-44, 10-15).

Indeed, in 111-4,4 innumerable references to the concept of supposition and to
propositions that must be distinguished can be found. In commentary on the
Sophistici Elenchi , the chapter dedicated to the third mode of equivocation (1,2,
§ 9) is the only part of this text where the concept of supposition is systematically
mentioned (other references to this concept in the text are scant and incidental).

Another kind of propositional ambiguity that is considered to be a case of
equivocation is that related to temporal propositions. 6 The fact that proposi-

6) Cf. De Sophisticis Elenchis, 166a, 2-7.

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

3 47

tions having a tensed copula must be distinguished is mentioned in different
parts of the Summa : the chapter concerning doubts that can be raised against
the notion of supposition in the first part (I, 72), the chapter concerning the
truth conditions of temporal propositions in the second part (II, 7), and the
chapter concerning equivocation in general (III-4, 2). In I, 72 and II, 7, Ock¬
ham accounts for the ambiguity of temporal propositions in terms of the
different things their subject terms can supposit for, which again shows the
conceptual affinity between the notion of supposition and the fallacy of equiv¬
ocation (in III-4, 2 he uses the term “stare pro ’ instead of “supponere pro').

And so any such proposition (“Socrates was a man”) is to be distinguished, in that such a
term can supposit for those who exist now or for those who existed in the past. Similarly, it
cannot supposit for those who will exist except in respect of a verb about the future and so
that proposition is to be distinguished, in that its term can supposit for those who exist or
those who will exist. Similarly for those who can be signified and are not, it cannot supposit
except in respect of a verb about the possible or the contingent, and so all such are to be
distinguished in that the subject can supposit for those which exist or for those which may
exist or may happen to exist. And so all such propositions as follows are to be distinguished:
‘Every man was white’, ‘Every white thing will be a man’, ‘Every white thing can be a man’,
‘It happens that every man runs’ (I, 72, 46-57)

And it should first be known that any proposition about the past and <any> about the
future is to be distinguished in that the subject can supposit for that which exists or that
which existed, if it is a proposition about the past, or for that which exists or for that which
will exist, if it is a proposition about the future. (II, 7, 3-8)

But if I were to say thus: “The white man was a man”, given that no man was ever white,
the subject does not stand for many things—that is, for those who were white or for those
who are white, because there never was such, but it is taken {denotatur) to stand for such in
this negative proposition, “The white man did not exist”. For it is taken to stand for such
only under disjunction, {sub disiunctione) (III-4, 2, 15-32)

At this stage of the analysis, De Rijk’s thesis appears to receive strong support
from the clear conceptual proximity between a theory of supposition that
focuses in cases of more than one different suppositum being possible, and the
concept of equivocation (in particular the third mode of equivocation). In
effect, equivocation can be explained as the semantic phenomenon according
to which a term signifies or supposits for different things, whereas the different
kinds of supposition can be justified by (the third mode of) equivocation.

Amphiboly

Prima facie , amphiboly differs from equivocation only with respect to the
length of the ambiguous phrase in question: equivocation concerns single

348

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

terms, whereas amphiboly concerns phrases composed of more than one term.
Just as with equivocation, there are three kinds of amphiboly.

About which it should be known first that, just as the fallacy of equivocation happens
because another term (dictio) can be taken in different ways, so the fallacy of amphiboly
happens because another expression ( oratio ) can be taken in different ways, without it being
the case that another term can first be taken in different ways; so that, just as the term is
multiple, so the whole sentence is multiple (111-4,5, 2-7).

In practice, complex expressions having more than one meaning are often
cases of syntactical ambiguity, that is, of the possibility of different parsings of
the phrase. So in this respect amphiboly differs from equivocation, which only
concerns semantic ambiguity. But Ockham also applies the concept of amphi¬
boly to propositions that feature semantic ambiguity, 7 in particular the second
mode of amphiboly (amphiboly by analogy). The latter is a very useful con¬
ceptual tool for Ockham; he uses it to interpret numerous cases of proposi¬
tions (in authoritative and holy texts) which, de virtute sermonis are false, but
whose author presumably meant something else (intentio auctoris ), namely a
secondary and derivative reading of the proposition (one that is true). 8

It should be known about the second mode of amphiboly that a proposition is multiple
according to the second mode of amphiboly when some proposition is properly, and from
its primary signification or imposition, taken in just one way, but improperly and secondar¬
ily it can be taken otherwise and have another sense. (II1-6, 6, 2-6)

... and in this way countless propositions in logic and metaphysics should be distinguished.
And likewise in natural philosophy, according to Aristotle’s opinion... (III-4, 6, 140-3)

In the chapters dedicated to amphiboly, there are very few explicit references
to the notion of supposition, 9 and that seems to indicate the absence of con¬
nection between the two notions. But if we consider a proposition that must
be distinguished on account of its subject having more than one possible

7) This is a genuine transformation of Aristotle’s original theory of fallacies.

8) Notice that this is a clear case of an ancient concept (amphiboly) being put to use in a different
context, where it acquires even greater importance.

9) Exception: ‘Quarta regula est quod quando accipitur modus cum dicto, ilia propositio est
distinguenda, eo quod potest denotari modus competere toti proposition! cuius est dictum, vel
proposition! de inesse in qua supponit pronomen demonstrativum vel nomen proprium alicuius
pro quo supponit subiectum respectu eiusdem praedicati mediante hoc verbo ‘est’ de praesenti.’
(III-4, 5, 67-72).

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

349

supposition (according to the third mode of equivocation), would it not be
correct to say that such a proposition is amphibolous? 10

And that there is some overlap between the different kinds of fallacies is
something recognized as unproblematic by Ockham. In particular, he men¬
tions the overlap between amphiboly and composition and division, 11 and
between equivocation and amphiboly.

It should be said that it does not matter much now whether the [given] case is considered
an instance of equivocation or of amphiboly. There can perhaps be amphiboly as much as
equivocation, and it does not present a problem that these fallacies should come together.
(111-4,7, 67-70)

And since such different senses can be had without any difference in punctuation, there is
amphiboly there, even if there should be composition and division there. (111-4,5, 110-3)

So it seems unproblematic that an occurrence of the third mode of equivoca¬
tion would be at the same time an occurrence of (the second mode of) amphi¬
boly. And if this is indeed the case, then there seems to be an important kinship
between Ockham’s theory of supposition and the fallacy of amphiboly, albeit
not explicitly formulated by him.

Division and Composition—Modal Propositions

As is well known, Ockham maintains that there are two kinds of modal prop¬
ositions, sine dicto and cum dicto , and that the latter is always ambiguous in
that it can be interpreted in the sense of division or in the sense of composi¬
tion (according to the two kinds of fallacies). (If interpreted in the sense of
division, it is equipollent in meaning to the corresponding modal proposition
sine dicto.)

A modal proposition said in the first way ( cum dicto) is always to be distinguished according
to composition and division. In the composite sense is always denoted that such a mode is
of the proposition of its dictum (de propositione illius dicti ). But the divided sense of such a
proposition is always equivalent to the proposition taken with the mode, without such a
dictum {sine dicto) (11,9, 12-25)

l0) Cf. Dutilh Novaes (2004b).

10 Ockham remarks that the kind of fallacy at stake can also be relative to a given language, such
that what in Latin seems to be a case of distinction according to amphiboly may be in Greek a
case of distinction according to division and composition (cf. III-4, 8, 43-38).

350

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

Again, Ockham does not explicitly relate the fallacies of composition and divi¬
sion to the concept of supposition in his analysis of modal propositions. But
just as much as in the case of temporal propositions, the ambiguity of modal
propositions cum dicto can be accounted for in terms of the supposition of its
terms: if u hominem esse animal est necessarium ’ is interpreted in the sense of
division, the subject supposits for actual men only, whereas the predicate sup-
posits for possible’ animals, but if it is interpreted in the sense of composition
both terms supposit with ampliation. 12 So if one views the fallacies of compo¬
sition and division in terms of scope (as Ockham seems to do), then their
relation with the notion of supposition is quite easily established. His analysis
of Aristotle’s example u possibile est sedentem ambulare ” (III-4, 8, 25-37) also
shows that composition and division are related to the scope given to the
modal expression in the proposition.

In any case, most of the examples discussed by Ockham in the chapter
dedicated to division and composition (III-4, chap. 8) are indeed cases of
modal propositions.

Fallacy ofFigure of Speech—Modes of Personal Supposition

A perhaps surprising point of contact between Ockham’s supposition theory
and his theory of fallacies concerns the fallacy figurae dictionis and the modes
of personal supposition. While the third mode of equivocation is used to jus¬
tify the different kinds of supposition, here the order of conceptual priority
seems to be inverted: the different modes of personal supposition are used as
conceptual tools to help identifying fallacies of figure of speech.

And in this mode consequences are badly formed where one mode of suppositing is changed
into another, as if it were argued thus: “Every man is an animal, therefore an animal is every
man.” And to make this example and others similar to it clear, it should first be known that
the fallacy of the figure of speech can be detected here, because of the fact that the same
term has one supposition in the antecedent and another in the consequent, although this is
not sufficient (III-4, 10, 212-8)

And it should be known that not every variation in supposition causes the fallacy of the
figure of speech... But in general there is the fallacy of the figure of speech when the
argument—at least when it is an enthymeme—goes from a term which is merely confused
in supposition to a term that is determinate in supposition, if no variation at all may be

l2) Notice though that the term “ampliation”, frequently used by other medieval authors to refer
to the supposition of a term for entities other than actual entities, is never used by Ockham
himself—see Priest and Read (1981).

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

351

made with regard to the other term and how it is disposed; although, along with this, there
is the fallacy of the consequent, as in the example proposed. (III-4, 10, 231-7).

It is necessary to know that the fallacy of the figure speech does not happen only to be
produced because of the different ways in which the same term supposits, but also because
of the different ways in which different terms supposit, in such a way that the other corre¬
sponding term is not changed. (III-4, 10, 262-3)

In fact, authors other than Ockham, such as Buridan, have provided an
account of legitimate inferential relations between categorical propositions on
the basis of the modes of personal supposition. 13 It is thus to be expected that
the modes of personal supposition would also serve to identify certain illegiti¬
mate inferences, which according to Ockham fall under the fallacy of figure
of speech. The same procedure is used in the commentary on the Sophistici
Elenchi II, 7 § 2, where various fallacies of figure of speech are cast in terms of
illegitimate moves of ascent and descent involving terms having different
kinds of personal supposition. 14

But it must be added that this fallacy is broader than what can be accounted
for in terms of the modes of personal supposition: only a small subset of falla¬
cies figurae dictionis seems to fall in this category.

Propositio est distinguenda

The doctrine that many propositions must be distinguished is a fundamental
aspect of Ockhams logic and of his treatment of authoritative and holy texts—
propositio est distinguenda is probably one of the most frequently occurring
phrases in the whole Summa. Most notably, it occurs with remarkable fre¬
quency in the final chapters of part I, in relation to Ockhams theory of
supposition, and in III-4, in particular in the chapters in which language-
dependent fallacies are analysed.

Such are to be distinguished in the same way... And so with many such. (I, 63, 22 and 27)

Secondly, it should be noted that any proposition in which such an equivocal term is placed
should always be distinguished de virtute sermonis, in that it can be taken in this way or that
[way], and this is so whether it is true in one sense and false in another, or whether it is false
in both senses or true in both. (III-4, 2, 62-5)

13) Karger (1993); Dutilh Novaes (2004a).

14) But Ockham adds that ‘non omnis diversitas suppositionis causat figuram dictionis
{Exp. Elench. p. 190, 22-23). Indeed, in some cases there can be a valid inferential relation
between propositions having the same terms in different personal suppositions (namely the valid
inferential patterns described in Karger (1993).

352

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

Therefore, we can not only conclude that ‘distinguishing propositions’ is a
crucial aspect of Ockham’s thought, but also that his theory of supposition as
much as his theory of fallacies are two of the main tools employed to accom¬
plish this task.

... It is worthwhile for philosophy to acquire the art that lets one know in how many ways
a multiple <proposition> is said. But this art (of the Sophistici Elenchi ) is of this sort so far
as the fallacies in speech are concerned, since in them the argument fails because something
is multiple. (Commentary on Sophistici Elenchi II, 1, 16-20)

(Preliminary) Conclusions

Given the criteria for comparison between two theories adopted here,
Ockham’s theories of supposition and of fallacies seem to bear significant
resemblance to each other: the apparatus seems very similar, at least in many
aspects, and there appears to be an important common goal, that of “distin¬
guishing propositions”.

An issue still to be clarified is whether the supposition apparatus can per¬
form all the legitimate distinctions of propositions, or whether there are prop¬
ositions which must be distinguished, but not in virtue of their terms having
more than one kind of supposition. It seems to me that there are more propo¬
sitions to be distinguished than there are propositions that can be distin¬
guished by using the supposition apparatus, and in this sense supposition
theory alone would not be sufficient for the performance of this task; this
might be one reason why Ockham still needs a theory of fallacies, to attempt
to cover the gap between what supposition theory can do and all the distinc¬
tions of propositions that must be made.

Differences

In sum, the number of aspects in which supposition theory and fallacy theory
appear to meet is rather impressive. But there are at least just as many aspects in
which there seems to be no point of contact between the two theories. In what
follows I outline some of them, starting with elements of the theory of fallacies
that have no apparent counterpart in supposition theory, and then vice-versa.

Fallacies: The Fallacy of Accent and Most Fallacies of Figure of Speech

The fallacy of accent obviously has no relation to the notion of supposition,
even more so since, as Ockham says it himself, it is a fallacy that only occurs

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

353

in spoken language; insofar as supposition is a property of spoken as well as of
written and mental terms, it should have no connection with a property exclu¬
sive to spoken language.

As for the fallacy of figure of speech, unlike the other language-dependent
fallacies (equivocation, amphiboly, composition and division), this fallacy
does not concern propositions that may rightly receive two interpretations
(which are “ distinguenda')\ rather, it seems to concern propositions that are
easily wrongly interpreted.

After the fallacies in speech, in which there are some sentences that are multiple, we should
speak of the fallacy of the figure of speech, which does not arise from the fact that any sen¬
tence is multiple, but arises rather from the likenesses of some sentences. (III-4, 10, 1-5)

More often than not, what is at stake in fallacies of figure of speech are syncate-
goremata and other syntactical properties of propositions, and not the supposi¬
tion of its terms. The exception to this rule is the use of the modes of personal
supposition to identify some of these fallacies, mentioned in the previous sec¬
tion, but this is not surprising insofar as the modes of personal supposition are
determined primarily by syntactic and syncategorematic aspects of the propo¬
sition (word-order and the presence or absence of quantifying terms).

Fallacies: Fallacies “extra dictionem”

The distinction between fallacies that are language-dependent and those that
are not is particularly relevant in the context of Ockham’s philosophy, as it is
directly related to the issue of mental language. 15 According to Ockham, lan¬
guage-dependent fallacies are those that occur only in written and spoken
language, whereas those that are not language-dependent also occur in mental
language. 16 In practice, Ockham’s views are not free of difficulties, in particu¬
lar concerning supposition in mental language and the third mode of equivo¬
cation. 17 It is rather telling that the notion of supposition is virtually absent
from the chapters concerning language-independent fallacies; this suggests
that supposition theory as it is formulated by Ockham is above all a machinery
to be applied to written and spoken language only (although mental terms
also supposit, but perhaps in different ways).

15) Cf. Knuuttila (2003).

16) Cf. Summa III-4, 1,22 40.

17) Cf Dutilh Novaes (2004b), Spade (1980a).

354

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

Among the fallacies extra dictionem , some concern primarily propositions
within arguments. Belonging to this group are the fallacy of the consequent,
the fallacy of begging the question; the fallacy of non-cause as cause; and the
fallacy of many questions. 18 But since the phenomena provoking these falla¬
cies belong to the propositional level, there is at most an indirect relation to
the supposition of the terms in question. From this group, only some fallacies
of the consequent are explicitly treated in terms of the personal supposition of
terms, 19 in a way similar to what had been done for some fallacies of figure of
speech.

As for the other three fallacies, fallacy of accident, fallacy secundum quid et
simpliciter and fallacy of misconception of refutation, only the fallacy of acci¬
dent seems to have some conceptual proximity to the notion of supposition—
it concerns the failure to identify the proper logical form of a proposition;
therefore, it can at times correspond to the failure of identifying the right sup-
posita of the terms in question. But Ockham does not mention the notion of
supposition in the chapter dedicated to this fallacy, which is all the more sur¬
prising since it would seem that at least some fallacies of accident could be
treated with the supposition machinery. 20 Rather, this kind of fallacy is mainly
treated within the framework of syllogistics in Summa 111-4,11.

In fact, except for the aforementioned case of fallacies of the consequent, I
am not aware of any reference to the notion of supposition in any of the seven
chapters (III-4, 11-17) dedicated to fallacies extra dictionem. Similarly, virtu¬
ally no reference to this notion in made in connection to language-independ¬
ent fallacies in the commentary on the Sophistici Elenchi . 21

Supposition: The Uses of the Notion of Supposition That Seem to Have No
Counterpart in the Theory of Fallacies

Even if one accepts the historical accuracy of De Rijk’s thesis, it is evident that
there are aspects of supposition theory—in any case in the fourteenth century—

18) It must be noted though that the fallacy of many questions can be caused by the terms being
understood in different ways (suppositing for different things, for example). In this sense this
fallacy has a conceptual proximity with the ‘ propositio est distinguenda issue, but this connection
is not explicitly made in Ockhams text.

19) Cf. Summa III-4. 12.

20) The fallacy of accident is the most important of the language-independent fallacies for Ock¬
ham, and these brief remarks are not intended to exhaust its relevance.

21) Exception: a brief reference to supposition in connection with the fallacy secundum quid et
simpliciter , Commentary on the Sophistici Elenchi , 268.

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

355

that fall entirely out of the scope of fallacy theory, being thus genuine medieval
innovations. When and how this separation took place is an interesting his¬
torical question which I shall not address now. I will simply mention some of
Ockhams uses of the notion of supposition that are not reducible to the fal¬
lacy framework.

Supposition: Modes of Personal Supposition

The modes of personal supposition may be helpful in identifying some falla¬
cies figurae dictionis , but the basic fallacies are of no use in defining the modes
of personal supposition. Indeed, this fragment of the theory does not concern
ambiguities—the terrain par excellence of fallacies.

Ockham could have said that the moves of ascent and descent that are not
allowed are fallacies, but he does not.

Personal supposition is merely confused when the common term supposits personally and
descent to a disjunction of singulars is not possible. (I, 70, 44-6)

So it seems evident that not only with respect to the conceptual apparatus but
also with respect to the goals to be achieved, no point of contact exists between
the fragment of supposition theory concerning the modes of personal suppo¬
sition and the theory of fallacies (besides the one already mentioned in the
previous section).

Supposition: The Theory of Truth-Conditions of Propositions

As is well known, one of the main uses of the concept of supposition is the
definition of the truth-conditions of categorical propositions recursively in
terms of the truth of demonstrative propositions of the form “This is a \ But
in the chapters in which this fragment of the theory is expounded, no refer¬
ence is made to fallacies, and neither can I see any obvious conceptual connec¬
tion. Ockham formulates the truth-conditions of propositions on the basis of
the supposition of its terms: in general, if the subject supposits for at least
some of the supposita of the predicate, then the proposition is true.

But what is sufficient for the truth of such an indefinite <proposition>, if it is indefinite? It
should be said that it suffices for the truth of such a proposition that, if it is affirmative, the
subject and predicate supposit for the same things, or, if it is negative, that they do not

supposit for the same thing—just as this suffices for the truth of a singular proposition_

(II, 3, 72-6)

356

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

Again, this fragment of the theory does not concern ambiguities and propo¬
sitions that must be distinguished, and thus it appears to have nothing in
common with the fallacies framework.

Different Objects and Different Goals: A Theory of Fallacies Concerns Arguments
(Chains of Propositions); a Theory of Supposition Concerns Terms and Propositions

A straightforward way to appreciate the significant dissimilarity between Ock¬
hams supposition theory and his theory of fallacies is the simple examination
of the examples he presents. While discussing the notion of supposition, the
examples brought up by Ockham are, with no exception that I know of, cases
of propositions taken as such. By contrast, in the exposition of his theory of
fallacies, virtually all the examples provided are of propositions within the con¬
text of arguments. In the latter context, Ockham is not interested in proposi¬
tions taken in isolation, but rather in propositions insofar as they are a part of
a chain of propositions (usually two or three) that constitute an argument.
Granted, in both cases the issue of propositions that must be distinguished
plays a central role, but within the theory of fallacies its importance is related
to the validity or invalidity of the argument in which the propositions are
found (and that obviously depends on the interpretation they are to receive).

In sum, it can be said that the objects of each theory are different: a theory
of fallacies concerns arguments and a theory of supposition concerns terms
and propositions. Naturally, insofar as arguments are composed of proposi¬
tions, there is a relation between these objects; just as much as propositions are
composed of terms, and therefore a property of terms such as supposition can
be used to distinguish propositions, arguments are composed of propositions,
and therefore properties of propositions must be taken into account to assess
the validity of arguments. But the fundamental difference between the three
logical levels—term level, propositional level and argument level—determines
the different scopes of each theory. In practice, each level corresponds to one
part of the Summa —first part for terms, second part for propositions, and
third part for arguments, and it is thus not surprising that the analysis of the
notion of supposition is to be found in the first two parts, whereas the theory
of fallacies is expounded in the third part.

A Theory of Fallacies Focuses on Error ; a Theory of Supposition Does Not

I have claimed that distinguishing propositions is a general common goal of
Ockhams theories of supposition and of fallacies. But the distinguishing of
propositions has different purposes within each theory.

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

357

In the theory of supposition, propositions must be distinguished within the
general framework of the interpretation of authoritative texts; the general aim
of supposition theory seems to me to be that of an “algorithmic hermeneu¬
tics”, the determination of the range of legitimate readings of a given proposi¬
tion. 22 In this sense, the emphasis of supposition theory is on, so to say, the
positive aspect of logic, in particular the definition of the moves that are
allowed in the context of propositional interpretation. The treatment of inter¬
pretations that are not allowed—of error—is at most an indirect one.

In the theory of fallacies, propositions must be distinguished insofar as the
failure in doing so can cause an illegitimate inference to be drawn. A theory of
fallacies is not a theory of valid arguments; rather, it is essentially a theory of
how invalid arguments can have the appearance of being valid. Hence, the
emphasis of a theory of fallacies appears to be on error, namely the identification
of inferential moves that are not allowed.

In sum, while within supposition theory the purpose of distinguishing
propositions is that of expanding the range of (interpretational) authorized
moves, in the theory of fallacies the same distinction of propositions concerns
the reduction of the range of (argumentative) authorized moves.

That being said, it is worth noting that one of the aspects in which Ock¬
ham’s theory of fallacies seems to deviate from the Aristotelian classical theory
is precisely its use for distinguishing propositions, which in Ockham’s case
appears to become one of the actual goals of his theory of fallacies, along with
the recognition of illegitimate inferences. Therefore, Ockham’s treatment of
propositions which are ‘distinguenda or ‘multiplex within the framework of
fallacies really seems to be an original application of this theory, one which
brings it closer to supposition theory.

It should be known that philosophy is acquired in two ways—through discovery and
through being taught. And because philosophy is acquired by being taught either by listen¬
ing to someone else or by looking at the books of others, and often such people, when they
or talking or writing, use multiple [i.e. ambiguous] propositions, which are true in one
sense and false in another sense, the learned needs to know the multiplicity of such propo¬
sitions so as not to take the false for the true or vice versa. And so it is useful to be able to
disambiguate sentences that are fallacious because of their multiplicity. (Commentary on
the Sophistici elenchi II, 1, 22-9)

22) For details of my (rather idiosyncratic) views on supposition theory, see Dutilh Novaes
(2007), chap. 1, and Dutilh Novaes (forthcoming).

358

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

Conclusions

Thus, there seems to be an overlap of conceptual apparatus and of goals (con¬
cerning propositions that must be distinguished) in Ockhams theories of sup¬
position theory and of fallacies. But the respective conceptual apparatuses also
present substantial dissimilarities, and their general goals are different: the the¬
ory of supposition defines the range of legitimate readings of (some) proposi¬
tions, while the theory of fallacies outlines illegitimate inferential patterns. So
all in all, Ockhams theory of supposition and his theory of fallacies seem indeed
to be two rather dissimilar theories, in spite of some (significant) overlap. 23

Hence, theories of supposition are better seen as an addition to the general
logical framework that medieval authors had inherited from ancient times,
rather than the replacement of an ancient theory by a medieval one. Indeed,
these theories and fallacy theories had different tasks to fulfil, and in this sense
both had their place in fourteenth century logic. One can formulate the fol¬
lowing hypothesis: a theory of supposition was badly needed in a culture of
textual commentary like the medieval one, but correctness of reasoning also
played a central role therein, and therefore a theory of incorrect but seemingly
correct reasoning was certainly also necessary. Thus, there was room for ancient
logical theories as well as for new ones.

Concerning these two theories, it seems that the ancient tradition was not
absorbed and replaced by the medieval new findings, but rather that they co¬
existed in the later medieval period (even though it is manifest that medieval
authors did not simply copy the ancient traditions even in the cases of conti¬
nuity). The task to be fulfilled by a theory of fallacies—the identification of
illegitimate arguments—could not be fulfilled by supposition theory, whereas
the tasks to be fulfilled by the concept of supposition also went beyond what
could be done only with the fallacy apparatus. Moreover, as I have suggested,
supposition theory alone could not account for all the propositions that must
be distinguished, so Ockhams theory of fallacies was perhaps also intended to
handle “ multiplex ” propositions that supposition theory could not handle.

This analysis also outlines what I consider to be one of the crucial traits of
Ockham’s general logical and philosophical enterprise: how he apparently
embraces the Aristotelian tradition and authorities in general, but in fact gives

23) The conclusion of the mutual conceptual independence of these theories may serve as
motivation to revisit De Rijks historical thesis, even though only extensive historical analysis
could serve to disprove it—and extensive historical analysis is not the aim of the present
investigation.

C. Dutilh Novaes / Vivarium 45 (2007) 343-359

359

them his own twist. In practice, this can be seen in his emphasis on “ propositio
est distinguenda , and in the way he uses, interprets and transforms the theory
of fallacies for his own philosophical purposes.

These remarks lead to one important issue that I have not dealt with in the
present discussion: the extent to which Ockhams theory of fallacies modifies
the original Aristotelian theory of fallacies. Roughly, it seems that there are a
couple of interesting modifications, even though in essence it is still the same
theory (the same 13 fallacies). But a more thorough comparison of the two
texts (Aristotle’s De Sophisticis Elenchis and part III-4 of Ockham’s Summa)
would be required to deal with this issue, alongside with a deeper analysis of
the commentary on the Sophistici Elenchi.

Finally, it is worth noting that theories of fallacies in fact outlived theories of
supposition; theories of supposition sort of faded away in the sixteenth cen¬
tury, 24 but theories of fallacies continued to play a central role in logic well
into the nineteenth century. 25

24) Cf. Ashworth (1982).

25) ‘Before the middle of the nineteenth century, textbooks of logic commonly taught the stu¬
dent how to check the validity of an argument (say in English) by showing that it has one of a
number of standard forms, or by paraphrasing it into such a form. The standard forms were
syntactic and/or semantic forms of argument in English. The process was hazardous: semantic
forms are almost by definition not visible on the surface, and there is no purely syntactic form
that guarantees validity of an argument. For this reason most of the old textbooks had a long
section on “fallacies”—ways in which an invalid argument may seem to be valid.’ (Hodges 2001).
See also Hamblin (1970).

BRILL

Vivarium 45 (2007) 360-373

VIVA

RIUM

www.brill.nl/viv

Richard Billingham’s Speculum puerorum ,
Some Medieval Commentaries and Aristotle

Egbert P. Bos 1

Leiden University

Abstract

In the history of medieval semantics, supposition theory is important especially in the
twelfth and thirteenth centuries. In this theory the emphasis is on the term, whose proper¬
ties one tries to determine. In the fourteenth century the focus is on the proposition, of
which a term having supposition is a part. The idea is to analyse propositions in order to
determine their truth (probare). The Speculum puerorum written by Richard Billingham
was the standard textbook for this approach. It was very influential in Europe. The theory
of the probatio propositionis was meant to solve problems both in (empirically oriented)
scientific propositions such as used by the Oxford Calculators, and theological proposi¬
tions, especially those about the Trinity. The book is original, concise, but not clear in every
respect. Studying medieval commentaries may help us to understand Richards book.
In the present paper three commentaries are presented. The commentators discussed
problems about the status of Richard’s book, and about its doctrine: what is the relation
between probatio and truth, what is the relation between probatio and supposition, what
exactly are mediate and immediate terms (e.g. is the pronoun this’ mediate or immedi¬
ate?). The commentators sometimes criticize Richard. For example, one of them argues,
against Billingham, that the verb ‘can’ ampliates its subject term and is therefore mediate.

Keywords

analysis of the proposition, supposition, mediate and immediate terms, resolvable and
componible terms

The subject of this paper, Richard Billingham’s Speculum puerorum , or Specu¬
lum iuvenum , or Terminus est in quern, or De probationibus propositionum was
composed about 1340, or even as late as 1330 in Oxford. 2 Apart from the

0 I would like to thank Stephen Read for correcting my English.
2) For more details of his life, see De Rijk (1982b), *17*.

DOI: 10.1163/156853407X217821

E. P. Bos / Vivarium 45 (2007) 360-373

361

Mirror , Richard wrote tracts on suppositions, ampliations, appellations, con¬
sequences, 3 obligations, insolubles; he also composed a collection of sophisms
with the title Conclusiones , a work on the sophism Numquid scire sit credere , a
short tract on the significate of a proposition, 4 5 a tract on the problem Utrum
idem Sortes est Sortem esse , and a tract On Compounded and Divided Sense?

Three Commentaries on the “Mirror”

In my analysis of the Mirror , I have used three medieval commentaries. The
first two have not been edited and investigated yet. The third, the sophistria
from Prague, is a part of a large work that I have published recently. 6 The com¬
mentaries differ in kind:

1. A short anonymous commentary, preserved in MS Munich, CLM 4383,
ff. 107va-112rb, dating from the late fourteenth century, perhaps the early
fifteenth century. 7 8 The commentator first gives some of the opening words of
a specific section of the Speculum , and adds his comments, which vary from
130 to 500 words. It consists of notes, starting with note’ (‘ sciendum). Short
though it is, it contains interesting criticisms on Billingham’s Mirror? This
commentary seems to have been preserved in one manuscript only.

2. A large commentary, mentioned, as Schum indicates in the catalogue of
the Amploniana collection of Erfurt, in an Erfurt catalogue of 1412. There it
is ascribed to Henricus Coesfeld. 9 Probably this Henry of Coesfeld is the same
as the one who wrote some theological tracts, prayers, preserved in a manu¬
script in Ljubljana. 10

3) Recently, Stephanie Weber (2003) has nicely edited Richards Consequences together with the
edition of a medieval commentary on the basis of MS. Toledo, Archivo del Catedral\ Cab., 94-27.

4) Cf. De Rijk (1976).

5) The most recent survey is to be found Weber (2003), 319-321. Weber gives references to
earlier surveys.

6) Bos (2005).

7) Henceforth, I refer to this commentary as ‘the Munich commentary’.

8) For instance about Billingham’s rejection of ampliation before a verb, as in ‘Adam is dead’. It
is said that Richard cannot explain the truth of this proposition [f. 108ra-b]: Sciendum: magister
in presenti loco non [f. 108rb] tenetur quia fundat se in illo quod terminus ampliativus non
potest ampliare ante se. Sed hoc est contra communem viam.

9) Schum (1887), 795.

10) Narodna in univerzitetna knijzinica v Ljubljani 148 (Kos 48). Home page: http://www.nuk.
uni-lj.si/vstop.egi.

362

E. P. Bos / Vivarium 45 (2007) 360-373

As far as I know, the commentary is preserved in two manuscripts, viz. MS
Krakow, Biblioteka Jagiellonica 2103, ff. 43v-80v, and Erfurt, Amplon. , Q243,
ff. lra-51rb. The text starts by quoting Proverbs (14,7): ‘ Vade contra virum
stultum ! n In the introduction Henry warns the reader that ignorance of logic
leads to sins and heresies.

The differences between the two manuscripts are great: the one has sections
omitted by the other. There are many variant readings, which burdens the
apparatus criticus. Here, too, one finds criticisms on Billingham’s tract.

3. A sophistria tract from Prague, ca. 1400, preserved in MS Cracow, Bibl.
Jag. 686, ff. lra-79rb. 12 The tract (consisting of two parts) is a specimen of the
art of sophistria . As the author himself says, the tract is intended to help stu¬
dents to avoid problems concerning fallacies in arguments and ambiguities of
words. A master of arts or aspirant master dictated it to young students (who
elsewhere, though not in this tract, are called l sophiste\ perhaps because of
their participation in disputes). The purpose of this approach is, as Ashworth
says, to prepare the student for the required exercises in sophistria.

The sophistria consists of a series of questions which presuppose knowledge
of various handbooks, such as those by Peter of Spain, Thomas of Cleves, John
Buridan, Marsilius of Inghen, Thomas Manleveld, Richard Billingham and
others. Though the sophistria is unfinished, and leaves many problems raised
in the tract, unanswered, the discussions are very interesting, even more than
in the ‘regular commentaries, and reveal many difficulties involved in Rich¬
ard’s Mirror.

The Present Paper

In the present paper I first make a comparison between Aristotle and Billing¬
ham with respect to their conception of logic and its relation to the science
of the time in which each lived. Then I shall compare the theory of supposi¬
tion and that of the probationes terminorum on the basis of what the com¬
mentators say. Then I shall discuss some problems of Billingham’s tract itself
with the help of some hitherto unknown medieval commentaries. These
commentaries were probably written in the fifty years after the publication
of the Mirror.

Next follows the definition of the subject matter of Billingham’s Mirror.
What is it about, especially in the view of his commentators? Finally I shall

10 I intend to publish (parts) of these commentaries.

12) On this sophistria , see the introduction to my edition, esp. ch. 4, and Ashworth (2006).

E. P. Bos / Vivarium 45 (2007) 360-373

363

discuss problems connected with the two of the four kinds of terms to which
different kinds of proof correspond. I shall comment on resolvable and com-
ponible terms, and on the corresponding kinds of proof, which can be com¬
pared to induction and deduction respectively.

Billingham and Aristotle

Billingham is a member of the long tradition of western logic and semantics,
which starts with Plato and Aristotle. Especially in his Posterior Analytics, Aris¬
totle is interested in necessary propositions. As elsewhere in his Organon Aris¬
totle’s focus is logical, but his interest in general and necessary propositions
may very well be said to correspond to his interest in general natures in reality.
His conception of knowledge was very strict. 13

Aristotle not only presented a theory of knowledge, he also practised science
itself, for instance, physics, astronomy and meteorology. The problem in the
sciences, however, is that the individual things studied do not always behave in
the same way. They are generated and decease, they grow and change. A scien¬
tist admits that those individual things only behave as they do in a majority of
cases (‘ ut inpluribus , as the Latin runs). This leads to the question how general
and necessary propositions about them can be formulated. Billingham’s logic
seems to be a new approach in tune with a new science in the fourteenth cen¬
tury which focuses on changing things and describes them mathematically.

In the Middle Ages new approaches in logic were developed, such as the
theory of supposition, and that of the probationes propositionum\ Billingham
follows the latter approach. Still, logicians frequently referred to Aristotle as
their main source of inspiration, whatever approach they chose. Thus Billing¬
ham made frequent references to the Philosopher. The commentator Henry of
Coesfeld labelled the Mirror as ‘ satis nova\ but this only in the sense of newly,
or recently, invented’. Nevertheless, Henry also says that Billingham’s logic is
more commonly taken as belonging to the logica vetus. ]A

13) Cf. Ackrill (1981), 36 and De Rijk (2002), II, 3 just to mention two prominent works.

14) Henry of Coesfeld, Commentary [MS E: f. 2va; K. f. 44v]: Tunc probatur propositio: hec
scientia est de partibus argumentationis tamquam de suis obiectis, igitur est pars logice veteris. Con¬
sequents tenet per notabile premissum, et assumptum patet, quia est de propositionibus.

Et confirmatur, quia subiectum huius scientie signifkat propositiones, et similiter partes subi-
ective significant propositiones, igitur etcetera. Assumptum patebit ex secundo dubio. Verum
tamen est quod, si quis vellet scientiam istam vocare ‘novam’, quia noviter est inventa, tunc ista
esset satis nova in ordine ad novam logicam. Sed primus modus est communior loquendi apud
philosophos.

364

E. P Bos / Vivarium 45 (2007) 360-373

Billingham’s Mirror is about the probationes propositionum (proofs of prop¬
ositions’), which words are also used as a title of the tract. It is a tract on logic,
which is based on the proof of general terms, which may help us to analyse
propositions. We shall see below that this kind of approach is not the same as
the theory of supposition. 15

Billingham’s tract is on logic, and gives a formal analysis. One could say, I
think, that a theory on the probationes propositionum is especially fruitful in
the analysis of scientific propositions, especially because individual terms play
a large part in exponible, componible and resolvable propositions (on these
terms, see below). 16 The matter of propositions, i.e. what they are about, may
interest us. Richard’s uses propositions about individual things with changing
properties, for instance, propositions about Sortes being bigger than Plato,
about Sortes beginning to be white, etc. 17 Fourteenth century logicians focused
on the analysis of those kinds of propositions. They wished to get to grips with
propositions on individual things and their changing properties, so not with
necessary propositions, such as ‘all men are mortal’.

Generally speaking fourteenth century philosophy took evidentia and certi -
tudo as the notions in which they were especially interested, and the theory
of the probationes propositionum seems to be especially supportive of this inter¬
est. 18 From 1333 onward we find this new kind of logic in England, particu¬
larly suited to analyse propositions of physics. In that period, there was a new
physics arising in England. The extent to which this new physics determined
the new approach to logic, is an interesting question, as Courtenay says. 19

Supposition of Terms and Proofs (or: Analyses) of Propositions

It is important to distinguish the two notions of probationes of propositions
and suppositio of terms.

Let us first discuss the theory of the probationes propositionum. As has been
said, Billingham’s Speculum uses the latter approach. The theory is intended to
analyse complex propositions about individual things, about ‘more’ and ‘less’,
about ‘to begin’ and ‘to cease’ and so on, and to distinguish their meanings.
Syncategorematic terms such as ‘all’ and ‘than’ play an important part. The

15) See below, p. 365.

16) See below, p. 369.

17) See also below, p. 365.
,8) De Rijk (1975), 99.

I9) Courtenay (1987), 240.

£ P. Bos / Vivarium 45 (2007) 360-373

365

theory of the proofs of propositions is one of the new tools to interpret
textbooks.

If one wishes to prove a proposition with an analysable term such as ‘man’,
one looks for the inferior terms under that term, for instance ‘ hoc under
‘homo ’. This analysis of a term is expressed in an analysis of a proposition, and
thus results in a syllogism. For instance: ‘homo currit’\\&s an analysable term,
viz. ‘homo. It is analysed thus: hoc currit, et hoc est homo, ergo homo currit ’.
There may also be a probatio of a proposition with what is called an exponible
term, for instance ‘ incipit\ Now the problematic term ‘ incipit ’ is analysed, its
meaning is made clearer: hoc incipit esse: hoc nunc est et immediate ante hoc non
fuit ita quod hoc et, igitur hoc incipit esse. The analysis of this kind of term helps
us understand syncategorematic terms. The argument is about something indi¬
vidual 0 hoc r ), and one can easily see how the proposition is verified. The refer¬
ence to inferior terms points to the empirical bent of this theory, I believe.

Here it is clear how the theory of the probationes is useful for the analysis of
propositions of natural science. This can be shown, according to Curtis Wilson,
for the case of Heytesbury’s Regule and Sophismata . 20 Though not mentioning
Billingham, John Murdoch’s study on propositional analysis (he discusses
William of Heytesbury) points in the same direction. Propositions with ‘begins’
and ‘ceases’ can very well be analysed by the probationes literature. 21

One of the novelties in medieval semantics is supposition theory. Especially
since De Rijk’s pioneering works of 1962-1967, tracts about the properties of
terms such as supposition, ampliation, appellation, have been edited and studied.

The nature of the theory of supposition is not easy to determine. Still, it
seems worthwhile trying, because this may help also us to understand the
theory of the proofs of propositions mentioned above. I shall not enter here
into discussions about the precise nature of the theory of supposition. Recently,
stimulating research has been done by Jack Zupko for Buridan’s logic, 22 and by
Catarina Dutilh Novaes for Ockham’s. 23

20) Wilson (1945), 12-13.

21) Murdoch (1979).

22) Zupko (2003), ch. 4.

23) In her recent book Catarina Dutilh Novaes discusses different interpretations of the theory.
According to her, it does not explain the mechanisms of reference nor determine the referent.
She concludes, quite convincingly, that the theory of supposition can be best viewed as a theory
of what she calls ‘algorithmic hermeneutics’. It is a theory to help interpret propositions. It must
be remembered that the theory is formal, but semantic, syntactic and pragmatic considerations
play a part. A purely formal analysis without regard to outward reality did not exist in the Middle
Ages: see Dutilh Novaes (2005), esp. § 1.6.

366

E. P. Bos / Vivarium 45 (2007) 360-373

The tract on the probationespropositionum is not about the truth as such of
such propositions. Supposition and truth are not the same, at least not in, for
instance, Ockham’s and Buridan’s views, but supposition makes clear to what
things the one who utters the proposition commits himself. 24 In the theory of
proof, one tries to derive the proof or falsity of propositions from an analysis
of a term that can be analysed. 25 The term is called the mediate term. With the
help of the analysis one can determine the truth of propositions used in text¬
books much more easily, as the medieval commentators themselves say.

The First Subject of Billingham’s Mirror

What is Richard’s Mirror about? What is, to use medieval terms, its subiectum
primum (‘first subject’) or, in other words, its first formal object? What is its
difference from other logical tracts, especially those about supposition? How
is the proposition, of which it purports to give an analysis, or probatio
conceived—what is its status? What is the relation between the language in
which the proposition is construed, and the intellect and senses, which are—
in a certain respect—the basis? What is the relation between the notions of
truth, proof and knowledge?

Billingham seems to be interested in propositions with the following
subjects:

1. Propositions about contingent facts, especially about individuals, in both
affirmative and negative propositions. Even when discussing universal affir¬
mative propositions he is not interested in necessary, scientific conclusions,
for instance ‘all men are mortal’, which pretend to be true for all times. He
intends a contingent fact about all men, for instance ‘all men are white’.
Billingham is aware that one cannot prove such non-necessary proposi¬
tions in the traditional Aristotelian way. They can be explained and anal¬
ysed, so that their truth value can be determined. It may be added that
Aristotle also allows singular statements as premises 26 and non-universal
conclusions. 27

24) Zupko (2003), 61, and 310, n. 10.

25) Cf. DeRijk (1982b), *3*.

26) Kneale and Kneale (1971), 68.

27) Richard Billingham, Speculumpuerorum (in De Rijk, 1982b), II, 1, 23-44; II, 2, 23-44; II, 3,
26-46; 11,4, 25-45; II, 3, 12-27.

E. P. Bos / Vivarium 45 (2007) 360-373

367

2. Propositions about differences within a species, which are expressed by
propositions having a comparative or superlative form, for instance;
‘Socrates is stronger than Plato’. 28

3. Propositions about identity, or difference between individual things, for
instance ‘Socrates is different from Plato’. 29

4. Universal propositions about contingent facts (cf. 1), but with the explicit
mention of something excepted, or with underlining of the fact expressed in
contradistinction to other states of affairs (propositions with ‘only’, for
instance ‘only a man runs’). 30

3. Propositions about things as far as they are in time, i.e. things that are said
to exist before or after some moment of time. 31

6. Propositions as far as they are used in physics, where beginning and end of
change is at issue, for instance ‘this begins to be’. 32

7. Modal propositions with ‘officiable’, or: ‘functionalisable’ terms, such as ‘con¬
tingent’, ‘possible’, ‘impossible’; for instance, ‘it is contingent that you are’. 33

Apart from these specific kinds of propositions Billingham emphasizes the
importance of word order , especially in propositions with functionalisable
terms. In the final section of the tract he gives three rules about word order : 34

In general one could say that, according to Billingham, the primary and
most fundamental signification of a term is individual existing things. This
existential import is preserved when a term precedes a verb. When it comes
after a verb, the term denotes the same individual, but regardless its existence.

Billingham’s intention is to analyse what are called ‘probable’ propositions
(‘propositiones probabiles). The ‘term ‘ probabilis may have different meanings.

‘Probare can be used for demonstration in the strict sense, as synonymous with

28) Richard Billingham, Speculumpuerorum (in De Rijk, 1982b), II, 1,49-53; II, 2, 49-53; II, 3,
51-55; II, 4, 51-55; II, 5, 28-33.

29) Richard Billingham, Speculum puerorum (in De Rijk, 1982b), II, 1, 54; II, 2, 54; II, 3, 56;
II, 4, 57; II, 5, 34.

30) Richard Billingham, Speculum puerorum (in De Rijk, 1982b), II, 1, 54; II, 2, 54; II, 3, 56;
II, 4, 57; II, 5, 34.

31) Richard Billingham, Speculum puerorum (in De Rijk, 1982b), II, 1, 55; II, 2, 55; II, 3, 57;
II, 4, 58; II, 5, 34.

32) Richard Billingham, Speculum puerorum (in De Rijk, 1982b), II, 1, 56-61; II, 2, 56-61; II, 3,
58-75; II, 4, 59-65; II, 5, 35-36.

33) Richard Billingham, Speculum puerorum , (in De Rijk, 1982b), II, 1, 62-70; II, 2, 62-72; II, 3,
64-75; 11,4, 66-77; II, 5,37-41.

34) Here Richard Billingham discusses sophisms, which bring into discussion fundamentals of
medieval semantics. More on Billingham’s semantics in general in De Rijk (1982a), 167-184.

368

E. P. Bos / Vivarium 45 (2007) 360-373

‘ demonstrare. Here ‘pro bare has a general sense of putting to the test’. To analyse
may also be a species under it. Generally speaking, fourteenth century philosophy
directed their attention to the procedures for proving sentences of all types . 35

As I said above, truth and knowledge, especially scientific knowledge, on
the one hand, and analysis on the other, should be distinguished . 36 Billingham
does not explain the relation between truth , proof and knowledge . 37 In the
introduction to his commentary, in the fifth doubt, Henry of Coesfeld speaks
about the truth value of the propositions. Henry asks: what is the truth of a
proposition, and how is it to be distinguished from a true proposition? This
leads him to a rejection of a theory of complexe significabilia , at least of a too
realistic version of it.

He also rejects the thesis that the truth of a proposition is an inherent qual¬
ity of the proposition. His own opinion is that the truth of a proposition is a
rational entity about the proposition. Truth and falsity are in the soul, not in
things . 38 Henry seems to be in line with Ockham by identifying the truth with
the proposition, be it mental, written or spoken and by considering less accept¬
able the position that a spoken or written proposition is true in virtue of a
mental proposition, a view which comes closer to Buridan . 39

35) De Rijk (1982b), *4*.

36) Questions 4 and 5 of the second tract of the Prague sophistria are comments on §§ 26-29 of
Billingham’s Mirror (Bos, 2005, 357, 1. 9-360, 1. 3). The anonymous master elucidates these
notions. Here the question is whether for the truth of an indefinite proposition the truth of
its demonstratives is required (question 4). Other than in question 2 (Bos, p. 350, 1. 23-355, 1.
28), the question here is about truth , not about proof. A true proposition is one that can sometimes
be proved. To prove a proposition, as has been said, is to show its truth with the help of other propo¬
sitions. One can also prove the falsity of propositions, if‘proof’ is taken in a more general sense.

37) Henry of Coesfeld, Commentary [E: f. 3va; not found in MS K]: Nota quid sit ‘probatio’, et
quid sit ‘probare’. Unde in proposito probatio est in propositione veritatis ostensio, et probare est
veritatem alicuius propositionis ostendere. Similiter improbatio est falsitatis alicuius propositio¬
ns ostensio, et improbare est falsitatem alicuius propositionis ostendere, et hoc aliquando vere,
et aliquando apparenter.

38) Henry of Coesfeld, Commentary [E: f. 4va, K: f. 46v; I leave the variant readings aside]: Veri¬
tas propositionis est ens rationis circa compositionem vel divisionem dictionum consistens,
denominans ab extrinseco id quod per compositionem vel divisionem significatur. (...) [E: f. 5ra,
K: 46v]: Probatur sic, quia verum et falsum sunt in anima, et non in rebus, sicut bonum et
malum sunt in rebus, et non in anima.

39) Henry of Coesfeld, Commentary [E: f. 4va, K: f. 46v]: Sed quia opinio hec communiter repu-
tatur falsa et erronea, ideo contra earn primo pono talem propositionem: non sunt ponenda
complexe significabilia quemadmodum ipsi posuerunt. Et dico notanter ‘quemadmodum isti
posuerunt’ vocando quamlibet talem rem complexe significabile que complexe potest significari.
Tunc quelibet res mundi quantumcumque simplex est, complexe significabile est, quia quelibet
talis potest per orationem complexe significari. Unde simplicissima res que est Deus, potest

£ P. Bos / Vivarium 45 (2007) 360-373

369

In his sixth doubt Henry asks whether there is a single way by which each
proposition can be proved to be true or false. He concludes that a proposition
is true when it is in fact as it once was true to be signified to be true. He empha¬
sises ‘to be signified’ in order to avoid any realism; he wishes to emphasise the
part played by the intellect . 40 I would like to draw attention to Mrs Weber’s
result when investigating Billingham’s notion of inference. Here too is the
emphasis on the intellect. A consequence is valid when the consequent is
understood in the antecedent . 41

The analysis of a proposition leads to considering the relation between the
proposition and the senses and intellect. Richard is not clear on this aspect. In
the commentaries we find the conclusion that the senses especially play a part
in accepting immediate propositions as true, and the intellect in accepting a
priori truths (for instance the principle of contradiction ). 42

Kinds of Terms

Billingham distinguishes between resolvable, componible, exponible and func-
tionalisable (or ‘officiable’) terms. The definitions of the technical terms used
by Billingham are important for understanding his semantics . 43 The commen¬
tators discuss these kinds, not so much the definitions themselves, but the
examples contained under them.

Mediate and Immediate Terms

In his Mirror , Billingham starts by presenting a principal division, viz. between
mediate and immediate terms.

complexe significari per hanc orationem complexam ‘Deus est bonus’, ‘Deus est prima causa’,
nam Deum esse primam causam nihil est in re nisi Deus, et sic de aliis.

40) Henry of Coesfeld, Commentary [E: f. 6va, not found in MS K]: Quarum prima est ista quod
cuiuslibet propositionis cathegorice affirmative de presenti in qua non ponitur <terminus>
ampliativus, ad veritatem requiritur quod sic sit in re qualitercumque per earn precise et secun¬
dum eius totalem significationem significatur. Probatur inductive, quia ad hoc quod hec sit vera
‘Sortes currit’, requiritur quod sit in re sicut per eius totalem significationem significatur, scilicet
quod ‘Sortes currit’ sit aliqua res.

41) Weber (2003), xxv.

42) The anonymous Munich commentary says [f. 108rb]: Sciendum: ista propositio est nota per
sensum quia sic se habet quod ipsa negata non contingit earn probari per aliquod notius, et
quando habetur notitia terminorum illius propositioni<s> immediate, tunc statim intellectus
eius assentit per eius naturalem inclinationem.

43) For further discussion the reader is referred to Maieru (1982). See also Andrews (1993).

370

E. P. Bos / Vivarium 45 (2007) 360-373

Billingham defines ‘ terminus in general as that in which a proposition is
resolved, together with the verb ‘to be’ or ‘not to be’. Here he follows Aristotle.
On the basis of Boethius he adds that terms are the nouns and verbs out of
which a proposition is put together.

Terms are either immediate or mediate. According to Billingham, immedi¬
ate terms are: demonstrative pronouns, substantive verbs and ampliative
verbs . 44 They cannot be resolved. Mediate terms are resolvable, componible,
exponible or functionalisable. Immediate terms are therefore neither resolv¬
able, nor exponible, nor functionalisable.

There is a discussion in the commentaries about the difference between
immediate nouns and immediate verbs : 45 verbs are immediate because they have
no superior(s), but only inferior(s). An interesting example is the verb ‘to be’.
He implies, I believe, that, if‘being’ could be deduced on the basis of a middle
term, it would be a genus term and categorematic, not transcendent, as ‘being’
is traditionally conceived. Nouns , however, are immediate if they have no
terms under them. Such a noun is for instance ‘this’. They are connected with
individuals, which are irreducible.

Billingham says that the verbs ‘is’ and ‘can’ are both immediate. In a
joint article, Ashworth and Spade wonder how an unambiguous notion of
immediacy can cover this claim. The medieval commentators also found a
difficulty here. Both the anonymous Munich commentator and the anony¬
mous master of the Prague sophistria criticise Richard . 46 The anonymous of
Munich argues that ‘can ampliates its subject term and is therefore resolvable
and mediate. He adds that ‘can’ can at best be called ‘immediate’ in compari¬
son with other ampliative verbs.

44) A. Maieru (1982a), 405.

49) Cf. Henry of Coesfeld, Commentary [E: f. 8rb, K: f. 48r]: Et subdit consequenter auctor in
particula tertia dicens quod in hoc conveniunt omnes termini immediati quia non habent
medium per quod propositiones in quibus ponuntur, possunt ratione ipsorum terminus probari,
quia non habent aliquod [E: f.8va] medium inferius per quod possunt probari, ut sunt prono-
mina et sic alii, quia non habent aliquod superius sicut hoc verbum ‘est’ et consimilia. Et dicit
auctor noster etiam quod hoc verbum ‘potest’ est terminus immediatus qui non habet superius
per quod propositio in qua ponitur, probari.

46) The Prague Sophistria [f. 57va]: Nota: aliqui dicunt quod ly ‘est’ non est resolubile, nec potest,
ex eo quod potentia presupponit actum. Sed potius videtur esse oppositum , scilicet quod actus
presupponit potentiam, quia quicquid est in actu, hoc est in potentia, et non econverso, quia
Antichristus est in potentia et non in actu. Et ergo potest assignari alia ratio, quia ly ‘potest’ est
ampliativum, et ly ‘est’ non, ergo non est superius. Et hoc videtur velle Auctor quando dicit quod
ly ‘est’ est terminus immediatus, ideo quia non habet verbum superius non ampliativum {Sophis¬
tria ed. Bos, p. 317,1. 30-318,1. 3).

E. R Bos / Vivarium 45 (2007) 360-373

371

Resolvable and Componible Terms; Resolvable and Componible Proof

To understand the Speculum it is important to know the definitions of the
terms ‘ resolubilis and ‘ componibilis\ not in the least because Billingham himelf
does not give definitions in a proper sense.

Resolvable Terms and Resolvable Proof

First on resolvable terms and propositions. Billingham says 47 that a resolvable
term is a common term (noun or participle), that has an inferior term under
it. ‘Homo’is such a resolvable term. It has ‘hoc as a term under it . 48 Therefore
a proposition like ‘ homo currit’ can be resolved into ‘ hoc currit\ and ‘ hoc est
homo. That a term is resolvable is immediately manifest in the fact that a
proposition in which such a term occurs can be resolved. The notions ‘resolv¬
able term’ and ‘resolvable proposition are linked.

The notion ‘ resolution resembles ‘ inductio , because the conclusion is based
on a premise about an individual. The commentators are well aware of that.
One could argue, as is done in the commentaries, that, clearly, ‘every man
runs, an animal is a man, therefore an animal runs’ cannot be a resolvable
proof, because it is on the basis of an inferior term. This cannot be a proof,
however, for the major premise is less known than the conclusion, which is
contrary to the nature of proof as such. The point is that the argument pro¬
ceeds from what is better known, not that it is a proof.

The commentaries make clear that induction is not a resolvable proof. For
instance, in the sophistria , the master replies that in a resolvable proof, there is
only a single instant as basis: in induction, however, more, or even all singulars
are required. So he distinguishes between these two kinds of argument, and
emphasises that the former is not a real proofs The resolution of a proposition
is the basis of the expository syllogism . 50

47) Richard Billingham, Speculumpuerorum (De Rijk, 1982b) II, 2, 7; II, 3, 9; II, 4, 7; II, 5, 6.

48) MS K of Billinghams Speculum has ‘verb’, participle, adverb. Or is ‘proper noun intended?
Paul of Venice has adverbs instead of pronouns, for instance ‘somewhere’ (‘ alicubi')\ see Paul of
Venice (1979), 224,1. 19.

49) The Pragu z-sophistria [f. 59vb]: Respondetur quod in <argumento> resolutorio sumitur
solum unum inferius ad terminum ratione cuius fit probatio resolutoria; sed in inductione
sumuntur plura inferiora, scilicet plura, vel omnia, singularia (, Sophistria , ed. Bos, p. 330,
11. 14-16).

50) The Munich commentary [f. 108vb]: Hie magister declarat quoddam prius dictum, scilicet
quod ista regula predicta est fimdamentum syllogismi expositorii, quia negato syllogismo expos¬
itors syllogismus habet probari per istam regulam. Et syllogismus expositorius habet probari per

372

£ P. Bos / Vivarium 45 (2007) 360-373

Componible Terms and Componible Proof

Next on componible terms and componible proof This is a difficult subject,
because Richard only mentions the componible term in the recensio altera f
and does not define it. He gives an example: ‘ homo currit'is composed by the
premisses ‘omne animal currit\ et ( omnis homo est animal.

Componible proof is also a problem. Richard seems to distinguish between
a nominal componible proof and a verbal componible proof. So do the com¬
mentators, but from the texts one gets the impression that not everybody
understood Richard’s intention.

A componible nominal proof is a proof of a proposition on the basis of a
higher term, for instance every animal runs, man is an animal, therefore man
runs’. It resembles deduction, which can be put in a syllogistic form, whereas
resolvable proof, on the contrary, is on the basis of an inferior term, e.g. a
singular term, which by itself cannot be placed in a syllogistic form.

In the sophistria from Prague, an opponent adds: ‘on the basis of a term
which naturally is better known’, but the anonymous author replies that to
prove componibly is not identical to a deductive proof in the strict sense. It is
about a contingent statement . 52

“Biligare”: Something New?

Is Billingham’s Mirror just a continuation of ancient logic? Or is it something
new? As can be expected, the medievals themselves say that in the book some¬
thing was explained which could also be found before, i.e. in Aristotle. They
refer, in this case, to the last chapters of Aristotle’s Topics . 53

In the Munich commentary, a fictional opponent asks who is the author of
the Speculum. The commentator answers ‘Billingham’, and adds that Billing-
ham is said to have combined the old and new logic, which is expressed in the

conditiones syllogismi. Unde syllogismus expositorius est syllogismus cuius medium est termi¬
nus singularis singulariter et univoce tentus, et est duplex: affirmativus cuius conclusio est
affirmativa, et negativus cuius conclusio est negativa.

51) Richard Billingham, Speculum puerorum (De Rijk, 1982b) I, 2, 12; II, 3, 9; II, 4, 7; II, 3, 6.

52) The Prague sophistria [f. 59vb]: Nota: probare componibiliter est in probatione propositionis
ratione alicuius termini assumere terminum superiorem eidem. Et aliqui addunt ‘nobis naturaliter
notiorem’, sed de hoc dicitur sicud circa precedentem descriptionem (Bos, 2005, 331,11. 9-11).

53) The Munich commentary [f. 107va]: Ad primum respondetur: hoc complexum ‘propositio
probabilis respectu termini immediati in ea positi’, et propria passio eius est ‘verum vel falsum
probabile. Et licet in libro Topicorum etiam habetur, tamen non ratione primi termini mediati.

E. P. Bos / Vivarium 45 (2007) 360-373

373

word ‘ biligare'7 4 Perhaps he means that Billingham investigates both terms
(the subject of the old logic), and proofs (the subject of the new logic). This is
remarkable, for usually a tract could be placed in either of the two parts of
ancient logic. It is even more remarkable that Henry of Coesfeld says that
Billingham’s work is something new, although he is well aware that this is usu¬
ally thought otherwise, as has been said. 55

The theory of the probationes propositionum focuses on the analysis of an
analysable term used in a proposition. The Munich commentator adds that
the final cause of the book, i.e. that at which it aims, is the truth or falsity of
propositions. 56

The present volume discusses the roots of medieval logic. Which were the
roots of the tradition of the analysis of propositions, such as Billingham’s Spec¬
ulum presents? To Billingham’s commentator Henry of Coesfeld, the roots lay
in Aristotle. Aristotle did not explicitly discuss the theory of the probationes
propositionum , but that was not necessary, he adds, for the meaning of
officiable, resolvable and exponible terms was clear enough. 57 Be that as it may,
the roots of this theory can also be found in the search of certainty and evi¬
dence, which characterized fourteenth-century philosophy. In general one
might say that this approach to logic was influenced by the science arising in
fourteenth-century England, which is empirically and mathematically formu¬
lated, such as one may find in William of Heytesbury. The theory of the anal¬
ysis of a proposition in combination with the theory of the properties of terms
made logic richer.

54) The Munich commentary [f. 107va]: Ad secundum respondetur quod fuit reverendus magis-
ter Richardus Biligam Anglicus, de studio Uxoniensi. Et dicitur quasi ‘biligans’ duo, scilicet
novam et veterem loycam.

55) Cf. above, p. 363.

%) The Munich commentary [f. 107va]: Ad quintum respondetur quod communiter loquendo
de causis scientie huius libri, tunc sunt quattuor, scilicet causa materialis, subiectum et intentio
sunt unum et idem, de quibus habitum est; sed causa finalis seu utilitas est cognitio veritatis vel
falsitatis propositionum.

57) Henry of Coesfeld, Commentary [E: f. 2va; K: 45r]: Propter hoc non explicite tradidit istum
modum, quia credidit ilium satis planum esse ex significatione terminorum resolubilium,
officiabilium vel exponibilium.

BRILL

Vivarium 45 (2007) 374-392

VIVA

RIUM

www.brill.nl/viv

Bibliography

Abbreviations

Aristoteles latinus

BGPTMA

Beitrage zur Geschichte der Philosophie und Theologie des Mittelalters

CAG

Commentaria in Aristotelem graeca

CCCM

Corpus christianorum continuatio mediaeualis

CCSG

Corpus christianorum series graeca

CCSL

Corpus christianorum series latina

CIMAGL

Cahiers de Plnstitut du Moyen Age grec et latin

CLCAG

Corpus latinum commentariorum in Aristotelem graecorum

CSEL

Corpus scriptorum ecclesiasticorum latinorum

Grammatici graeca

Grammatici latini, ed. H. Keil, Hildersheim, 1964 (repr.)

Patrologia graeca , ed. J.-P. Migne

Patrologia latina , ed. J.-P. Migne

Sources chretiennes

Ackrill, J. L. (1981) Aristotle the Philosopher , Oxford.

Adam Wodeham (1990) Lectura secunda in librum primum Sententiarum , ed. G. Gal and
R. Wood, St. Bonaventure, N.Y.

Aertsen, J. A. (1991) ‘The Medieval Doctrine of the Transcendentals. The Current State of
Research’, Bulletin dephilosophie medievale 33, 130-147.

- (1996) Medieval Philosophy and the Transcendentals. The Case of Thomas Aquinas , Leiden,

New York and Cologne.

- (1999) ‘The Medieval Doctrine of the Transcendentals: New Literature’, Bulletin de phi¬
losophie medievale 41,107-121.

- (2000) ‘ Transcendens- Transcenden talis: The Genealogy of a Philosophical Term’ in Hamesse

and Steel (eds) Elaboration du vocabulaire philosophique au moyen age , Turnhout (SIEPM.
Rencontres de philosophie medievale, 8), 241-233.

- (2002) ‘“Res” as Transcendental: Its Introduction and Significance’ in G. Federici Vesco-

vini (ed) Leproblbne des transcendentaux du XIV e au XVII e siecles , Paris, 139-155.

Aertsen, J. A. and A. Speer (eds) (1998) Was ist Philosophie im Mittelalter?, Erfurt, Berlin and
New York (Miscellanea mediaevalia 26).

Alain of Lille (1981) Regulae caelestis iuris , ed. N.M. Haring, Archives dhistoire doctrinale et lit-
teraire du moyen age 48, 97-226.

Albert the Great (1889) Depraedicabilibus , ed. S. C. A. Borgnet, Opera omnia 1 , Paris.

- (1890) De Sophisticis Elenchis , ed. S. C. A. Borgnet, Opera omnia 2, Paris.

DOI: 10.1163/156853407X217830

Bibliography / Vivarium 45 (2007) 374-392

375

- (1893) In I Sententiarum, ed. S. C. A. Borgnet, Opera omnia 26, Paris.

Alcinous (1993) The Handbook of Platonism, trans. J. Dillon, Oxford.

Alexander Hales (1931) Glossa in quatuor libros sententiarum P. Lombardi , Florence.

Alfarabi (1932) Catalogo de las ciencias , ed. A. G. Palencia with Catalan translation, Madrid.

Allard, G.-H. (ed) (1986) Jean Scot ecrivain, Montreal and Paris.

Ammonius (1895) In Aristotelis Categorias Commentarius , ed. A. Busse, Berlin (CAG IV, 4).

- (1897) In Aristotelis De Interpretation Commentarius , ed. A. Busse, Berlin (CAG IV, 5).

Andrews, R. (1993) ‘Resoluble, Exponible and Officiable Terms in the Sophistria of Petrus Olai,
MS Uppsala, University Library, C 599’ in Read (1993), 3-33.

Anonymous (1964) Rhetorica ad Herennium. Incerti auctoris de ratione dicendi ad C. Herennium
lib. IV, ed. F. Marx, Leipzig (M. Tulli Ciceronis scripta quae manserunt omnia, fasc.l).

Anonymus ad Cuimnanum (1992) Expositio latinitatis , ed. B. Bischoff and B. Lofstedt, Turn-
hout (CCSL 133D).

Anonymus d'Orvillensis (1999) ed. S. Ebbesen, CIMAGL 70, 229-423.

Anselm of Canterbury (1946) Opera omnia , Vol. I, ed. F. S. Schmitt, Seckau, Rome and
Edinburgh.

Aristotle (1966) Categoriae et Liber de interpretation , ed. L. Minio-Paluello, Oxford.

- (1968) Aristoteles Latinus XXXIII Editio Altera. De Arte Poetica translatio Guillelmi de

Moerbeka. Accedunt Expositio Media Averrois sive 'Poetria Hermanno Alemanno interprete et
specimina translations Petri Leoni, ed. L. Minio-Paluello, Leiden.

- (1975) Aristoteles Latinus VI1-3. De Sophisticis Elenchis , ed. B. G. Dod, Leiden and Brussels.

- (1978) Aristoteles Latinus XXXI1-2. Rhetorica. Translatio Anonyma sive Vetus et Translatio

Guillelmi de Moerbeka, ed. B. Schneider, Leiden.

- (1984) The Complete Works of Aristotle: The Revised Oxford Translation, ed. J. Barnes, Princeton.

Ars Laureshamensis. Expositio in Donatum maiorem (1977) ed. B. Lofstedt, Turnhout (CCCM 40A).

Ashworth, E. J. (1982) ‘The Eclipse of Medieval Logic’ in Kretzmann, Kenny and Pinborg
(1982), 786-796.

- (1992) ‘Analogy and Equivocation in Thirteenth-Century Logic: Aquinas in Context’,

Mediaeval Studies 54, 94-135.

- (2006) ‘Logic Teaching at the University of Prague’, History of Universities 21,211—221.

- (forthcoming) ‘Equivocation and Divine Language in Some Theology Texts from the

Twelfth and Early Thirteenth Centuries’ in C. H. Kneepkens and H. A. G. Braakhuis (eds) 30
Years Logica Modernorum. Acts of the Symposium in Honor of Professor L. M. de Rijk, Turnhout
(Studia Artistarum 13).

Augustine (1958) Locutions in heptateuchum, ed. J. Fraipont, Turnhout (CCSL 33).

- (1962) De Doctrina Christiana, Turnhout (CCSL 32).

- (1968) De Trinitate, ed. W. J. Mountain, Turnhout (CCSL, 50A).

- (1975a) De Dialectica, trans. B. Darrell Jackson, ed. J. Pinborg, Dordrecht and Boston.

- (1975b) De diversis quaestionibus octoginta tribus, ed. A. Mutzenbacher, Turnhout

(CCSL, 44A).

- (1991) De Trinitate in Oeuvres de Saint Augustin, deuxieme serie, vols. 15 and 16, Turn¬
hout.

Averroes (1953) Commentarium magnum in Aristotelis De Anima, ed. F. S. Crawford, Cam¬
bridge, Mass.

Avicenna (1977-1983) Liber dephilosophia prima, ed. S. van Riet, Leiden.

Baine Harris, R. (ed) (1976) The Significance of Neoplatonism, Norfolk, Virginia.

376

Bibliography / Vivarium 45 (2007) 374-392

Barnes, J. (1990) The Toils of Scepticism, Cambridge.

Barwick, K. (1957) Probleme der stoischen Sprachlehre undRhetorik, Abhandlungen der Sdchsischen
Akademie der Wissenschaften zu Leipzig , Philologisch-historische Klasse 49/3, Berlin.

Bechtle, G. (2002), ‘Diairesis, Definition, Analysis, Synthesis: Betrachtungen zu Jamblichs
Skopos-Lehre und zur Interpretation des platonischen Sophistes (253dl-e5)’, Wiener Studien,
115, 175-218.

Bede (1863) Bedae Venerabilis Liber De Schematibus et Tropis , in Rhetores Latini Minores, ed.
C. Halm, Leipzig; repr. Dubuque, Iowa [no date].

Beierwaltes, W. (1986), ‘Language and Its Object. Reflexions on Eriugena’s Valuation of the
Function and Capacities of Language’, in Allard (1986), 209-228.

Biard, J. and I. Rosier-Catach (eds) (2003) La tradition medievale des Categories (XJI e -XV e
siecles), Louvain-la Neuve, Leuven and Paris (Philosophes medievaux 45).

Bobzien, S. (2002) ‘A Greek Parallel to Boethius’ De hypotheticis syllogismis’, Mnemosyne 55,
285-300.

Boethius (1847) Opera, PL 64.

- (1880) Commentarii in librum Aristotelis Peri hermeneias, pars posterior , ed. C. Meiser, Leipzig.

- (1906) In Isagoge Porphyrii commenta , ed. G. Schepss and S. Brandt, Vienna and Leipzig

(CSEL48).

- (1969) De hypotheticis syllogismis, ed. L. Obertello, Brescia (Istituto di Filosofiadell’Universita

di Parma, Logicalia 1).

- (1978) Boethius's De topicis dijferentiis, trans. E. Stump, Ithaca and London.

- (2000) De consolationephilosophiae; opuscula theologica, ed. C. Moreschini, Munich and Leipzig.

Bonfiglioli, S. (2006a) Agalma. Icone e simboli nel neoplatonismo tardoantico, Doctoral Thesis in
Semiotics, University of Bologna.

- (2006b) ‘Simboli o icone? I segni in limine della tradizione platonica in Avventure del

simbolo. Icone e simboli tra Antichita e Medioevo, ed. S. Bonfiglioli, Versus: Quademi di studi
semiotici 102, 27-64.

Bos, E. P. (2005) Logica modernorum in Prague about 1400. The Sophistria disputation ‘Quoniam
quatuor', Leiden.

Brehier, E. (1962) La theorie des incorporeb dans Tancien stoicisme, Paris.

Breton, S. (1981) ‘Symbole, image, prodige’, dialogue sur J. Trouillard (1981), Dialogues
d'histoire ancienne 7, 309-322.

Bruun, O. and L. Corti (eds) Les categories et leur histoire, Paris.

Buffa, M. F. (1982) ‘“Corpus” e “res” nella terminologia grammaticale latina’, Studi e ricerche
dell'Istituto di Latino 5, 7-28.

Burnett, C. (1995) ‘Adelard of Bath’s Doctrine on Universal and the Consolatio Philosophiae of
Boethius’, Didascalia, 1, 1-10.

Callus, Daniel A. (1948) ‘The ‘Tabulae super Originalia Patrum’ of Robert Kilwardby, O. P.,
d. 1279’ in Studia Mediaevalia in honorem admodum Reverendi Patris Raymundi Josephi Mar¬
tin, Ordinis Praedicatorum, Bruges, 243-270.

Cameron, M. (2004) ‘What’s in a Name? Students of William of Champeaux on the vox
significativa, Bochumer Philosophisches Jahrbuch fur Antike undMittelalter 9, 93-114.

Carchia, G. (1997) La favola dell'essere: commento al Sofista, Macerata.

Cardullo, L. (1985) II linguaggio del simbolo in Proclo , Catania.

Cavini, W. (1977) ‘Appunti sulla difFusione in occidente delle opere di Sesto Empirico’, Medioevo

3 , 1 - 20 .

Bibliography / Vivarium 45 (2007) 374-392

377

Cesalli, L. (2001) ‘Le realisme propositionnel de Walter Burley’, Archives d'histoire doctrinale
et litteraire du moyen age 68, 155-221.

- (2002) ‘Some Fourteenth Century Realist Theories of the Proposition. A Historical and

Speculative Study’ in Gill (2002), 83-118.

- (2005) ‘Le “pan-propositionnalisme” de Wyclif, Vivarium 43, 124-155.

- (2007) Le realisme propositionnel. Semantique et ontologie des propositions chez Jean Duns

Scot, Gauthier Burley, Richard Brinkley et Jean Wyclif, Paris.

Chiesa, C. (1991) Semiosis — signes — symboles. Introduction aux theories du signe linguistique de
Platon et d’ Aristote, Berne, Berlin and Paris.

- (1991a) ‘Le probleme du langage interieur chez les stoiciens’, Revue internationale de phi¬
losophy 178, 301-321.

- (1992) ‘Le probleme du langage interieur dans la philosophic antique de Platon a Por-

phyre’, Histoire epistemologie langage 14, 15-30.

Chydenius, J. (1960) The Theory of Medieval Symbolism, Helsinki.

Colish, Marcia (1985) The Stoic Tradition from Antiquity to the Early Middle Ages: 2. Stoicism in
Christian Latin Thought through the Sixth Century, Leiden, New York, Copenhagen and
Cologne.

Compendium logicae Porretanum (1983) Compendium logicae Porretanum, ed. S. Ebbesen, K. M.
Fredborg and L. O. Nielsen, CIMAGL 46, 1-113.

Conti, A. (1982) ‘Teoria degli universali e teoria della predicazione nel trattato De universalibus
di William Penbygull: discussione e difesa della posizione di Wyclif’, Medioevo 8, 137-166.

- (1993), ‘Logica intensionale e metafisica dell’essenza in John Wyclif’, Bullettino

dell'Istituto Storico Italiano per il Medio Evo e Archivio Muratoriano 99/1, 159-219.

- (1997) ‘Analogy and Formal Distinction: On the Logical Basis of Wyclif s Metaphysics’,

Medieval Philosophy and Theology 6, 133-165.

- (2000) ‘Significato e verita in Walter Burley’, Documenti e studi sulla tradizione filosofica

medievale 9, 317-350.

Coulter, J. A. (1976) The Literary Microcosm. Theories of Interpretation of the Later Neoplatonists,
Leiden.

Courtenay, W. (1987) Schools and Scholars in Fourteenth-Century England, Princeton.

- (1991) ‘Nominales and Nominalism in the Twelfth Century’, Lectionum varietates. Hommage

a Paul Vignaux (1904-1987), ed. J. Jolivet et al., Paris, 11-48.

Craemer-Ruegenberg, I. and A. Speer (eds) (1993) Scientia undars im Hochmittelalter und Spat-
mittelalter, Berlin and New York (Miscellanea mediaevalia 22).

Criscuolo, U. (1992) ‘Iconoclasmo bizantino e filosofia delle immagini divine nel Neopla-
tonismo’ in Gersh and Kannengiesser (1992), 83-102.

D’Onofrio, G. (1986a) Fons scientiae. La dialettica nell'Occidente tardo-antico, Naples.

- (1986b) ‘ Disputandi Disciplina. Procedes dialectiques et Logica vetus dans le langage

philosophique de Jean Scot’ in Allard (1986), 229-263.

Dahan, G. (1980) ‘Notes et textes sur la Poetique au Moyen Age’, Archives d’histoire doctrinale et
litteraire du moyen age 47, 171-239.

Dascal, M. et al. (eds) (1992) Sprachphilosophie /Philosophy of Language /Laphilosophie du langage.
Ein intemationales Handbuch zeitgenossischer Forschung/ An International Handbook of Contem¬
porary Research / Manuel international des recherches contemporaines, Berlin and New York.

De Libera, A. (1981) ‘A propos de quelques theories logiques de Maitre Eckhart: existe-t-il une tradi¬
tion medievale de la logique neoplatonicienne?’, Revue de theologie et de philosophie, 113, 1-24.

378

Bibliography / Vivarium 45 (2007) 574-392

- (1986) ‘Les Summulae dialectices de Roger Bacon: I—II. De termino, De enuntiatione’,

Archives d’histoire doctrinale et litteraire du moyen age 53, 139-289.

- (1987) ‘Les Summulae dialectices de Roger Bacon: III. De argumentatione’, Archives

d’histoire doctrinale et litteraire du moyen age 54, 171-278.

- (1994) ‘D’Avicenne a Averroes, et retour. Sur les sources arabes de la theorie scholastique

de l’un transcendental’, Arabic Science and Philosophy 4, 141-179.

- (1996) La querelle des universaux. De Platon a la fin du moyen dge, Paris.

- (1999a) L’art des generalites: theories de Pabstraction, Paris.

- (1999b) ‘Genese et structure des metaphysiques medievales’ in Narbonne and Langlois

(1999), 159-181.

- (2002), La reference vide. Theories de la proposition , Paris.

De Rijk, L. M. (1962-1967) Logica modemorum: A Contribution to the History of Early Terminist
Logic , Assen.

- (1966) ‘Some New Evidence of Twelfth-Century Logic: Alberic and the School of Mont

Ste Genevieve (Montani)’, Vivarium 4, 1-57.

- (1975) ‘The Place of Billinghams ‘Speculum puerorum’ in the 14th and 15th Century

Logical Tradition with the Edition of Some Alternative Tracts’, Studia Mediewistyczne 16,
99-153.

- (1976) ‘Richard Billinghams Works on Logic’, Vivarium 14, 121-138.

- (1977) ‘On Ancient and Medieval Semantics and Metaphysics’, Vivarium 15, 81-110.

- (1980) ‘The Semantical Impact of Abailard’s Solution to the Problem of Universal’ in

Thomas (1980), 139-151.

- (1982a) ‘Semantics in Richard Billingham and Johannes Venator’ in A. Maieru (ed) Eng¬
lish Logic in Italy in the 14th and 15th Centuries , Naples, 167-183.

- (1982b) Some 14th Century Tracts on the Probationes terminorum. Martin of Alnwick O. EM. ,

Richard Billingham, Edward Upton and Others , Nijmegen.

- (1986) Plato’s Sophist. A Philosophical Commentary, Amsterdam, Oxford and New York.

- (1994) ‘On Man’s Capability of Grasping the Truth’ in Craemer-Ruegenberg and Speer

(1994), 281-303.

- (2002) Aristotle, Semantics and Ontology, Leiden, Boston and Cologne (Philosophia

antiqua 90).

Denifle, H. (ed) (1889-1897) Chartularium Universitatisparisiensis, Paris.

Dijs, J. (1990) ‘Two Anonymous 12th-Century Tracts on Universals’, Vivarium, 28, 85-117.

Dillon, J. (1976) ‘Image, Symbol and Analogy: Three Basic Concepts of Neoplatonic Allegorical
Exegesis’ in Baine Harris (1976), 247-262.

Dionysius Thrax (1989) Techne, ed. J. Lallot, La grammaire de Denys le Thrace, Paris.

Dominic of Flanders (1967) Quaestiones super XII Libros Methaphysicorum, Venice 1499; repr.
Frankfurt.

Ducharme, S. (1940) ‘Note sur le transcendental “res” selon Saint Thomas’, Revue de I’Universite
d’Ottawa 9, 85-99.

Dutilh Novaes, C. (2004a) ‘The Buridanian Account of Inferential Relations between Dou¬
bly Quantified Propositions: A Proof of Soundness’, History and Philosophy of Logic 25,
215-234.

- (2004b) ‘Ockham on Supposition and Equivocation in Mental Language’. Proceedings of

the Society for Medieval Logic and Metaphysics 3, URL = http://www.fordham.edu/gsas/
phil/ldima/SMLM/index.htm.

Bibliography / Vivarium 45 (2007) 374-392

379

- (2005) Formalizations apres la lettre. Studies in Medieval Logic and Semantics, PhD disser¬
tation, Leiden University.

- (2007) Formalizing Medieval Logical Theories: Suppositio, Consequentia and Obligationes,

Berlin.

- (forthcoming) ‘An Intensional Interpretation of Ockhams Theory of Supposition’, Journal

of the History of Philosophy.

Ebbesen, S. (1976) ‘Anonymus Aurelianensis II, Aristotle, Alexander, Porphyry and Boethius.
Ancient Scholasticism and 12th century Western Europe’, CIMAGL 16, 1-128.

- (1979) ‘Anonymi Aurelianensis I Commentarium in Sophisticos Elenchos’, CIMAGL 34,

iii-xlviii, 1-200.

- (1981) Commentators and Commentaries on Aristotle s Sophistici Elenchi. A Study of Post-

Aristotelian Ancient and Medieval Writings on Fallacies , vols. I—III, Leiden (CLCAG VIL1-3)
[Vol. 1, pp. 133-170 repr. as ‘Porphyry’s Legacy to Logic: A Reconstruction’ in Sorabji (1990),
141-171].

- (1982) ‘Ancient Scholastic Logic as the Source of Medieval Scholastic Logic’ in Kretzmann,

Kenny and Pinborg (1982), 101-127.

- (1983) ‘The Odyssey of Semantics from the Stoa to Buridan’ in A. Eschbach and

J. Trabant (eds) History of Semiotics (=Foundations of Semiotics 7), Amsterdam, 67-85.

- (1987) ‘Boethius as an Aristotelian Scholar’ in Wiesner (1987), II, 286-311 [repr. with

minor changes in Sorabji (1990), 373-391].

- (1991) ‘[Review of] J. Magee, Boethius on Signification and Mind’, Vivarium 29, 150—

154 [corrigenda in Vivarium 31 (1993), 270 n. 1].

- (1997) ‘Texts on Equivocation. Part I. Ca.ll30-ca. 1270’, CIMAGL 67 , 127-199.

- (1998) ‘Texts on Equivocation. Part II. Ca. 1250-1310’, CIMAGL 68, 99-307.

- (1999) ‘Anonymus D’Orvillensis’ Commentary on Aristotle’s Categories’, CIMAGL 70,

229-423.

- (2003a) ‘Anonymus d’Orvillensis on the Categories in Biard and Rosier-Catach (2003),

347-363.

- (2003b) ‘Boethius on the Metaphysics of Words’ in Galonnier (2003), 257-275.

- (2003c) ‘Burley on Equivocation in His Companion to aTractatus Fallaciarum and in His

Questions on the Elenchi’, CIMAGL 74, 151-207.

- (2004) ‘Where Were the Stoics in the Late Middle Ages?’ in Strange and Zupko (2004),

108-131.

- (2005a) ‘Theories of Language in the Hellenistic Age and in the Twelfth and Thirteenth

Centuries’ in Frede and Inwood (2005), 299-319.

- (2005b) ‘Les categories au moyen age et au debut des temps modernes’ in Bruun and

Corti (2005), 245-274.

- (2005c) ‘Gualterus Burleus, Quaestiones super Sophisticos Elenchos 4-12. A revised edition’,

CIMAGL 76, 239-282.

Ebbesen, S. and Y. Iwakuma (1990) ‘Anonymus Parisiensis, Compendium Sophisticorum Elen-
chorum (ms. Paris BN 4720A)’, CIMAGL 60, 47-112.

Ebbesen, S. and I. Rosier-Catach (2000) ‘Robertus Anglicus on Peter of Spain’ in P. Perez-Ilzarbe
and I. Angelelli (eds) Medieval and Renaissance Logic in Spain, Hildesheim, Zurich and New
York, 61-95.

Elias (1900) Eliae (olim Davidis) In Aristotelis Categorias Commentarium, ed. A. Busse, Berlin
(CAG XVIII).

380

Bibliography / Vivarium 45 (2007) 374-392

Erchanbert (1948) Erchanberti Frisingensis tractatus super Donatum, ed. W. V. Clausen, Chicago.

Erismann, C. (2002) ‘ Generalis essentia. La theorie erigenienne de Xousia et le probleme des
universaux’, Archives d’histoire doctrinale et litteraire du moyen age, 69, 7-37.

- (2004) ‘ Processio id est multiplicatio. L’influence latine de l’ontologie de Porphyre: le cas de

Jean Scot Erigene’, Revue des Sciences Philosophiques et Theologiques 88, 401-460.

- (2003) l Un autre aristotelisme? La problematique metaphysique durant le haut Moyen age

latin. A propos d’Anselme, Monologion 27’, Quaestio. Annuario di storia della metafisica 5,
143-160.

Evrard of Bethune (1887) Graecismus , ed. J. Wrobel, Bratislava.

Fattori, M. and M. Bianchi (eds) (1982) Res. 111° Colloquio Internazionale del Lessico Intellettuale
Europeo, Rome (Lessico intellettuale europeo, 26).

Fehling, D. (1958) ‘[Review of] Karl Barwick, Probleme der Stoischen Sprachlehre’, Gottingische
Gelehrte Anzeigen 212, 161-173.

Firmin Le Ver (1994) Firmini Verris Dictionarius/ Dictionnaire latin-frangais de Firmin Le Ver,
ed. B. Merrilees and W. Edwards, Turnhout (CCCM. LLMA 1).

Fischer, B. (1912) De Augustini Disciplinarum libro qui est de Dialectica , dissertation, Jena.

Flavius Sosipater Charisius (1997), Artis grammaticae libri V, ed. K. Barwick, corr. F. Kiihnert,
Stuttgart and Leipzig.

Flury, P. (1982) ‘Res im antiken Latein. Allgemeine Oberlegungen und Beobachtungen zum
Sprachgebrauch einzelner Autoren’ in M. Fattori and M. Bianchi (eds) Res. 111° Colloquio
Internazionale del Lessico Intellettuale Europeo , Rome, 35-45.

Fodor, J. (2003) Hume Variations , Oxford.

Fodor, J. and E. Lepore (2002) The Compositionality Papers , Oxford.

Fredborg, K. M. (1976a) ‘Buridan’s Quaestiones super Rhetoricam Aristotelis in J. Pinborg (ed)
The Logic of John Buridan, Copenhagen, 47-59.

- (1976b) ‘The Commentaries on Cicero’s De inventione and Rhetorica ad Herennium by

William of Champeaux’, CIMAGL , 17, 1-39.

- (1977) ‘Tractatus glosarum Prisciani in MS Vat. lat. i486’, CIMAGL 21, 21-44.

- (1981) ‘Some Notes on the Grammar of William of Conches’, CIMAGL 37, 21-41.

- (1988) ‘Speculative Grammar’ in P. Dronke (ed) A History of Twelfth-Century Western Phi¬
losophy , Cambridge, 176-195.

- (1999) ‘Promisimus’, CIMAGL 70, 81-228.

- (2003) ‘Abelard on Rhetoric’ in C. Mews, C. Nederman and R. Thomson (eds) Rhetoric

and Renewal in the Latin West 1100-1540, Essays in Honour of John O. Ward, Turnhout, 54-80.

Fredborg, K. M., L. Nielsen and J. Pinborg, (1978) ‘An Unedited Part of Roger Bacon’s Opus
maius: De signis’, Traditio 34, 75-136.

Frede, D. and B. Inwood (eds) (2005) Language and Learning , Cambridge.

Gal, G. (1987) ‘Adam ofWodeham’s Question on the “Complexe significabile” as the Immediate
Object of Scientific Knowledge’, Franciscan Studies 37, 66-102.

Galonnier, A. (ed) (2003) Boece ou la chaine des savoirs , Louvain and Paris (Philosophes
medievaux 44).

Garlandus (1959) Dialectica , ed. L. M. de Rijk, Assen.

Gerard de Harderwyck (i486) Commentum super veterem artem Arestotelis secundum viam Alber-
tistarum, Cologne.

Gersh, S. (1978) From Iamblichus to Eriugena: an Investigation of the Prehistory and Evolution of
the Pseudo-Dionysian Tradition, Leiden.

Bibliography / Vivarium 45 (2007) 374-392

381

Gersh, S. and C. Kannengiesser (eds) (1992) Platonism in Late Antiquity, Notre Dame, Ind.,
83-102.

Ghellinck, J. de (1949) Patristique et Moyen Age. Etudes d'histoire litteraire et doctrinale, I: Les
recherches sur les origines du symbole des Apdtres, Gembloux, Brussels and Paris.

Gilbert of Poitiers (1966) The Commentaries on Boethius by Gilbert of Poitiers, ed. N. M. Haring,
Toronto (Studies and Texts 13).

Giles of Rome (1967a) Super libros Posteriorum Analyticorum, Venice 1488; repr. Frankfurt.

- (1967b) Super libros Elenchorum, Venice 1496; repr. Frankfurt.

- (1968) Commentaria in Rhetoricam Aristotelis, Venice 1315; repr. Frankfurt.

Gill, H. S. (ed) (2002) Signification in Language and Culture, Shimla.

Gilson, E. (1942) La philosophie au moyen age. Paris.

Gracia, J. J. E. (1984) Introduction to the Problem of Individuation in the Early Middle Ages,
Munich and Vienna.

Gracia, J. J. E. (1992) ‘The Transcendentals in the Middle Ages. An Introduction’, Topoi 11,
113-120.

Gregory of Rimini (1978-1987) Lectura super Primum et Secundum Sententiarum, ed. D. Trapp
and V. Marcolino, Berlin and New York (Spatmittelalter und Reformation. Texte und Unter-
suchungen 6-12).

Gregory the Great (1982) Registrum epistolarum, ed. D. Norgberg, Turnhout (CCSL 140A).

Gregory, T. (1958), ‘Sulla metafisica di Giovanni Scoto Eriugena, Giornale critico della filosofia
italiana 37, 319-332.

- (1977), ‘L’eschatologie de Jean Scot’, in AAW, Jean Scot Erigene et Phistoire de la phibso-

phie, Paris, 377-392.

Grellard, C. (2005) Croire et savoir. Les principes de la connaissance selon Nicolas dAutrecourt,
Paris.

- (forthcoming) ‘Nicholas of Autrecourt’s Skepticism’ in H. Lagerlund (ed) Skepticism in

Medieval and Renaissance Thought, Leiden.

Grondeux, A. (2002) ‘Materiaux pour une histoire de l’acception syntaxique de la synecdoque’,
Histoire Epistemobgie Langage 24/1, 119-174.

- (2003) ‘Corpus dicitur quidquid videtur et tangitur. Origines et enjeux d’une definition’,

Voces 14, 35-76.

Grondeux, A. and C. Jeudy (2001) ‘A propos de pus: sens medieval d’un mot antique’, Archivum
btinitatis medii aevi 59, 139-160.

Guillaume de Tailleur (2002) Dictionarius familiaris et compendiosus. Dictionnaire btirt-frangais
de Guilbume Le Tailbur, ed. Edwards, W. and B. Merrilees, Turnhout (CCCM. LLMA 3).

Hadot, P. (1980) ‘Sur divers sens du mot pragma dans la tradition philosophique grecque’ in
P. Aubenque (ed) Concepts et categories dans la pensee antique, Paris.

Halliwell, S. (2002) The Aesthetics of Mimesis. Ancient Texts and Modem Problems, Princeton and
Oxford.

Hamblin, C. L. (1970) Falbcies, London.

Hamesse, J. (1974), Les Auctoritates Aristotelis, un florilege medieval, Louvain and Paris.

- (1982) ‘ Res chez les auteurs philosophiques des XII e et XIII € siecles ou le passage de la

neutralite a la specificite, in M. Fattori and M. Bianchi (eds) Res. 111° Colbquio Intemazionale
del Lessico Intellettuab Europeo, Rome.

Heinzmann, R. (1974) Die Summa 'Co lligite fragmenta des Magister Hubertus (Clm 28779), ein
Beitrag zur theologischen Systembildung in der Scholastik , Munich.

382

Bibliography / Vivarium 45 (2007) 374-392

Henry Totting of Oyta (1512) Abreviatio (of Adam Wodeham’s Lectura), Venice.

Hodges, W. (2001) ‘Model Theory’ in The Stanford Encyclopedia of Philosophy (winter 2001 edition),
ed. E. N. Zalta, URL = <http://plato.stanford.edu/archives/win2001/entries/model-theory/>.

Hoenen, M. J. E M. (1993) Marsilius of Inghen. Divine Knowledge in Late Medieval Thought,
Leiden (Studies in the History of Christian Thought 50).

- (1997) 1 Propter dicta Augustini. Die metaphysische Bedeutung der Mittelalterlichen

Ideenlehre’, Recherches de Theologie et Philosophic Medievales 64, 245-262.

Hoffmann, T. (2002) Creatura intellecta. Die Ideen und Possibilien bei Duns Scotus mit Ausblick
auf Franz von Mayronis, Poncius undMastrius, Munster (BGPTMA n.f. 60).

Holtz, L. (1981) Donat et la tradition de Tenseignement grammatical. Etude sur T Ars Donati et sa
diffusion (iv'—ix e siecle) et edition critique , Paris.

Hopkins, J. and H. Richardson (2000) Complete Philosophical and Theological Treatises of Anselm
of Canterbury, Minneapolis.

Hiilser, K. (1992) ‘Stoische Sprachphilosophie’, in Dascal (1992), 17-34.

Hunt, R. W. (1941-1943) ‘Studies on Priscian in the Eleventh and Twelfth Centuries. I. Petrus
Helias and his Predecessors’, Medieval and Renaissances Studies 1, 2 (repr. in R. W. Hunt,
Collected Papers on the History of Grammar in the Middle Ages, ed. G. L. Bursill-Hall, Amster¬
dam, 1980, 194-231).

Iamblichus (1937) De vita Pythagorica, ed. L. Deubner, Leipzig (editio corr.: ed. U. Klein, Stutt¬
gart, 1975).

- (1966) Les mysteres dEgypte, ed. fi. Des Places, Paris.

Incerti auctores (1977) Quaestiones super Sophisticos Elenchos, ed. S. Ebbesen, Copenhagen
(Corpus philosophorum danicorum medii aevi 7).

Invisibilia Dei (1973) ed. N. M. Haring, Recherches de theologie ancienne et medievale 40,
104-146.

Isidore of Seville (1906) Etymologiae, ed. W. M. Lindsay, Oxford.

Iwakuma, Y. (1992) ‘ Vocales or early nominalists’, Traditio, 47, 37-110.

- (1993) ‘Introductions dialecticae artis secundum magistrum G. Paganellum’, CIMAGL

63, 45-114.

- (1995) ‘Nominalia’, Didascalia 1, 47-88.

- (1996) ‘The Realism of Anselm and His Contemporaries’ in D. Luscombe and G. R.

Evans (eds) Anselm: Aosta, Bee and Canterbury, Manchester, 120-135.

- (1997) ‘Enuntiabilia in Twelfth-Century Logic and Theology’ in Marmo (1997), 19-35.

- (1999) ‘Pierre Abelard et Guillaume de Champeaux dans les premieres annees du

XII' siecle: une etude preliminaire’ in J. Biard (ed) Langage, sciences , philosophic au XITsiecle,
Paris, 92-123.

- (2003a) ‘William of Champeaux and the Introductions’ in H. Braakhuis and C. H.

Kneepkens (eds) Aristotle’s Peri Hermeneias in the Latin Middle Ages, Essays on the Commentary
Tradition, Groningen and Haren, 1-30.

- (2003b) ‘William of Champeaux on Aristotle’s Categories in J. Biard and I. Rosier-Catach

(eds) La tradition medievale des Categories (XII e -XV e siecles), Louvain and Paris, 313-328.

Jacobi, K. (2003) ‘Nomina transcendentia. Untersuchungen von Logikern des 12. Jahrhunderts
(iber transkategoriale Term’ in Pickave (2003) 23-36.

Jacobi, K., C. Strub and P. King (1996) ‘From intellectus verus/falsus to the dictum propositio-
nis: The Semantics of Peter Abelard and His Circle’, Vivarium 34, 15-40.

James of Viterbo (1983) Jacobi de Viterbio O. E. S. A. Quaestiones de Divinis Praedicamentis 1-X,
ed. E. Ypma, Rome (Corpus scriptorum augustinianorum 5 pt. 1).

Bibliography / Vivarium 45 (2007) 374-392

383

Jeauneau, E. (1972) ‘Allegoria, Mysterium, Sacramentum, Symbolum’ in E. Jeauneau (ed) Jean
Scot , Commentaire sur I'Evangile de Jean , Paris, 397-402 (SC 180).

Jeudy, C. (1977) ‘Israel le grammairien et la tradition manuscrite du commentaire de Remi
d’Auxerre a 1 ’Ars minor de Donat’, A Gustavo Vinay = Studi Medievali 3a serie, 18, II.

John Buridan (1998) Johannes Buridanus. Summulae de Suppositionibus , ed. R. van der Lecq,
Nijmegen (Artistarium 10-4).

- (1999) Summulae de Demonstrationibus, ed. L. M. de Rijk, Nijmegen (Artistarium 10-8).

- (2001) Summulae de Dialectica, trans. G. Klima, New Haven and London.

John Damascene (1935) De Fide orthodoxa. Versions of Burgundio and Cerbanus , ed. E. M. Buy-
taert, St. Bonaventure, N.Y. (Engl, transl. in Salmond 1898).

John Duns Scotus (1639) Quaestiones in primum et secundum librum Perihermeneias Aristotelis et
In duos libros Perihermeneias operis secundi quaestiones , in Opera omnia I, ed. Wadding, Lyon.

- (2004), Quaestiones in Libros Perihermenias Aristotelis; Quaestiones Super Librum Elencho-

rum Aristotelis; Theoremata , ed. R. Andrews et al., St. Bonaventure, N.Y.; Washington, D.C.
(Opera philosophica 2).

John Scottus Eriugena (1939) Annotationes in Marcianum , ed. C. E. Lutz, Cambridge, Mass.

- (1972) Commentaire sur I’Evangile de Jean , ed. E. Jeauneau, Paris (SC, 180).

- (1975) Expositio in Hierarchiam coelestem, ed. J. Barbet, Turnhout (CCCM 31).

- (1996-2003), Periphyseon. ed. E. Jeauneau, Turnhout (CCCM 161-165).

John Wyclif (1893-1899) Tractatus de logica , ed. M. H. Dziewicki, London.

- (1909) De ente, ed. M. H. Dziewicki, London.

- (1985a) De universalibus, ed. I. J. Mueller, Oxford.

- (1985b) On Universal , trans. A. Kenny, Oxford.

Jolivet, J. (1975) ‘Notes de lexicographie abelardiennes’ in Pierre Abelard. Pierre le Venerable ,
Paris (Colloques internationaux du CNRS 546), 531-545.

- (1966) ‘Quelques cas de “platonisme grammatical” du VII C au XII C siecle’, in Melanges

offerts a Rene Crozet , Poitiers, 93-99 (repr. in Jolivet (1987), 71-77).

- (1981) ‘Non-realisme et platonisme chez Abelard. Essai d’interpretation’ in Abelard en son

temps , Paris, 175-195 (repr. in Jolivet (1987), 257-277).

- (1987) Aspects de la pensee medievale: Abelard\ Doctrines du langage , Paris.

- (1991) ‘Donnees sur Guillaume de Champeaux, dialecticien et theologien in J. Longere

(ed) L’abbayeparisienne de Saint-Victor au Moyen Age, Paris and Turnhout, 235-251.

John of Salisbury (1991) Metalogicon , ed. J.B. Hall and K.S.B. Keats-Rohan, Turnhout (CCCM 98).

Julian of Toledo (1973) Ars grammatica, poetica, rhetorica, ed. M. A. H. Maestre Yenes, Toledo.

Kaluza, Z. (2003) ‘La notion de matiere et son evolution dans la doctrine wyclifienne’ in M.-T.
Beonio-Brocchieri Fumagalli and S. Simonetta (eds) John Wyclif. Logica, Politica, Teologia ,
Florence, 113-151.

Karamanolis, G. E. (2006), Plato and Aristotle in Agreement? Platonists on Aristotle from Antiochus
to Porphyry , Oxford.

Karger, E. (1993) ‘A Theory of Immediate Inferences Contained in Buridan’s Logic’ in K. Jacobi
(ed) Argumentationstheorie, Leiden, 407-430.

- (1996) ‘Mental Sentences according to Burley and the Early Ockham’, Vivarium 34, 192-230.

- (1999) ‘Walter Burley’s Realism’, Vivarium 37, 24-40.

Kavanagh, C. (2005) ‘The Influence of Maximus the Confessor on Eriugena’s Treatment of
Aristotle’s Categories’, American Catholic Philosophical Quarterly 79, 567-596.

Kelly, L. (2002) The Mirror of Grammar: Theology, Philosophy and the Modistae , Amsterdam and
Philadelphia (Studies in the History of the Language Sciences 101).

384

Bibliography / Vivarium 45 (2007) 374-392

Kenny, A. (1985) Wyclif, Oxford.

- (1986) ‘The Realism of the De universalibus', in A. Kenny (ed) Wyclif in His Times , Oxford,

17-29.

Kerenyi, K. (1962) ‘ArAAMA, EIKQN, EIAQAON’ in Demitizzazione e immagine, Archivio di
Filosofia , 161-171.

King, P. (1982) Peter Abelard and the Problem of Universal, Princeton (unpublished dissertation).

- (1987) ‘Jean Buridan Philosophy of Science’, Studies in History and Philosophy of Science

18, 109-132.

- (2004) ‘The Metaphysics of Peter Abelard’ in J. E. Brower and K. Guilfoy (eds) The Cam¬
bridge Companion to Abelard, Cambridge, 65-125.

- (forthcoming.) ‘Abelard on Mental’, American Catholic Philosophical Quarterly.

Kirwan, C. (2001) ‘Augustine’s Philosophy of Language’ in N. Kretzmann and E. Stump (eds)
The Cambridge Companion to Augustine, Cambridge, 186-204.

Kneale, W. C. and M. Kneale (1962) The Development of Logic, Oxford.

Kneepkens, C. H. (1978) ‘Master Guido and His View on Government: On Twelfth Century
Linguistic Thought’, Vivarium 16, 108-141.

- (1992) ‘Nominalism and Grammatical Theory in the Late Eleventh and Early Twelfth

Centuries. An Explorative Study’, Vivarium 30(1), 34-50.

- (1993) ‘Orleans 266 and the Sophismata Collections: Master Joscelin of Soissons and the

Infinite Words in the Early Twelfth Century’ in Read (1993), 64-85.

- (1995) ‘The Priscianic Tradition’ in S. Ebbesen (ed) Sprachtheorien in Spdtantike und

Mittelalter (= Geschichte der Sprachtheorie, P. Schmitter III), Tubingen.

- (1997) ‘Please, Don’t Call Me Peter: I Am an Enuntiabile, Not a Thing. A Note on the

Enuntiabile and the Proper Noun’ in Marmo (1997), 83-98.

Knuuttila, S. (2003) ‘The Question of the Validity of Logic in Late Medieval Thought’ in
R. Friedman and L. Nielsen (eds) The Medieval Heritage in Early Modern Metaphysics and
Modal Theory, Dordrecht, 121-142.

Kopp, C. (1982) ‘Ein kurzer Fehlschlufitraktat: Die Fallaciae breves (ad modum Oxoniae). London,
British Museum, Royal MSS 12 F XIX, 104rb-105vb’, Miscellanea Mediaevalia 15, 262-277.

- (1985) Die “Fallaciae ad modum Oxoniae”: Ein Fehlschlufitraktat aus dem 13. Jahrhundert ,

inaugural-dissertation, Cologne.

Kretzmann, N. (ed) (1988) Meaning and Inference in Medieval Philosophy, Dordrecht.

Kretzmann, N., A. Kenny and J. Pinborg (eds) (1982) The Cambridge History of Later Medieval
Philosophy: From the Rediscovery of Aristotle to the Disintegration of Scholasticism. 1100-1600,
Cambridge.

Ladner, G. B. (1979) ‘Medieval and Modern Understanding of Symbolism: A Comparison’,
Speculum 54(2), 223-256.

Lallot, J. (1989) La grammaire de Denys le Thrace , Paris.

Lambert of Auxerre (Lagny) (1971) Logica (Summa Lamberti), ed. F. Alessio, Florence.

Law, V. (2003) The History of Linguistics in Europe from Plato to 1600, Cambridge.

Law, V. and I. Sluiter (1998) Dionysius Thrax and the Techne Grammatike, Munster (The Henry
Sweet Society Studies in the History of Linguistic Ideas 1).

Lenz, M. (2003) Mentale Sdtze: Wilhelm von Ockhams Thesen zur Sprachlichkeit des Denkens, Stuttgart.

- (2004) ‘Oratio mentalis und Mentalesisch: Ein spatmittelalterlicher Blick auf die gegen-

wartige Philosophic des Geistes’ in J. Aertsen and M. Pickave (eds) Herbst des Mittelalters?
Fragen zur Bewertung des 14. und 15. Jahrhunderts, Berlin and New York (Miscellanea mediae¬
valia 31), 105-130.

Bibliography / Vivarium 45 (2007) 374-392

385

- (2005) ‘Peculiar Perfection: Peter Abelard on Propositional Attitudes’, Journal of the His¬
tory of Philosophy 43(4), 377-386.

Lewry, P. O. (1978) Robert Kilwardbys Writings on the ‘Logica Vetus Studied with Regard to Their
Teaching and Method , unpublished dissertation, Oxford.

Lloyd, A. C. (1956) ‘Neoplatonic Logic and Aristotelian Logic’, Phronesis 1, 58-72; 146-160.

- (1971) “Grammar and Metaphysics in the Stoa” in A. A. Long (ed) Problems in Stoicism,

London,58-74.

- (1982), ‘Procession and Division in Proclus’ in H. J. Blumenthal and A. C. Lloyd (eds)

Soul and the Structure of Being in the Late Neoplatonism: Syrianus, Proclus and Simplicius,
Liverpool, 18-45.

Lofstedt, B. (1982) Ars Ambrosiana, Commentum anonymum in Donatipartes maiores , Turnhout
(CCSL 33C).

Long, A. A. (1971) “Language and Thought in Stoicism” in A. A. Long (ed) Problems in
Stoicism , London, 75-113.

- (1981) ‘Aristotle and the History of Greek Scepticism’ in D. J. O’Meara (ed) Studies in

Aristotle , Washington, D.C., 79-106.

Long, A. and D. Sedley (1987) The Hellenistic Philosophers, Vol. 1: Translation of the Principal
Sources with Philosophical Commentary , Cambridge.

Luhtala, A. (2000a) ‘Early Medieval Commentary on Priscian’s Institutiones Grammaticae’,
Cl MAGL 71, 115-88.

- (2000b) On the Origin of Syntactical Description in Stoic Logic , Munster.

- (2000c) ‘Glosses Based on Eriugena’s Priscian Commentary’, Miscellanea Bibliothecae

Apostolicae VaticanaeT , 199-213.

Luna, C. (1990) ‘Commentaire et notes a la traduction de Simplicius, Commentaire sur les
Categories', translation directed by I. Hadot, Fascicule III, Leiden.

Maestrelli, C. A. (1976) ‘Riflessi linguistici della simbologia nell’Alto Medio Evo’ in Simboli e
simbologia nellAlto Medioevo , I, Spoleto, 789-811 (Proceedings of the 23rd Settimana di
Studio del Centro Italiano di Studi sull’Alto Medioevo).

MacDonald, S. (1991) ‘The Metaphysics of Goodness and the Doctrine of the Transcendentals’
in S. MacDonald (ed) Being and Goodness. The Concept of the Good in Metaphysics and Philo¬
sophical Theology, Ithaca, N.Y. and London, 31-55.

- (1992) ‘Goodness as Transcendental: The Early Thirteenth-Century Recovery of an Aris¬
totelian Idea’, Topoi 11, 173-186.

Magee, J. (1989) Boethius on Signification and Mind, Leiden, New York, Copenhagen and
Cologne (Philosophia antiqua 52).

Maieru, A. (1982a) Terminologia logica della tarda scolastica , Rome (Lessico intelletuale europeo 8).

- (ed) (1982b) English Logic in Italy in the 14th and 15th Centuries, Naples.

- (1999) ‘Littera e allegoria’ in Storiografia epoesia , Rome, 137-150 (Nuovi studi storici 35).

Maieru, A. and L. Valente (eds) (2004) Medieval Theories of Assertive and Non-assertive Language,
Rome.

Maioli, B. (1974) Gli universali. Storia antologica del problema da So crate al XII secolo, Rome.

Mann, W. (1993) ‘[Review of] J. Magee, Boethius on Signification and Mind’, Gnomon 95,
498-502.

Marenbon, J. (1980), ‘John Scottus and the Categoriae decern’ in W. Beierwaltes (ed) Eriugena.
Studien zu seinen Quellen, Heidelberg, 117-134.

- (1992) ‘Abelard, ewand Unity’, Topoi 11, 149-158.

- (1997a) The Philosophy of Peter Abelard, Cambridge.

386

Bibliography / Vivarium 45 (2007) 374-392

- (1997b) ‘The Platonisms of Peter Abelard’ in L. G. Benakis (ed) Neoplatonisme etphiloso¬
phic medievale , Turnhout, 109-129.

- (2000) Aristotelian Logic, Platonism and the Context of Early Medieval Philosophy in the

West, Aldershot and Burlington.

- (2004a) ‘Dicta, Assertion and Speech Acts: Abelard and Some Modern Interpreters’ in

Maieru and Valente (2004), 59-80.

- (2004b) ‘Life, Milieu and Intellectual Context’ in J. Brower and K. Guilfoy (eds) The

Cambridge Companion to Abelard, Cambridge, 13-44.

- (2005) ‘Les Categories au debut du Moyen age’ in Bruun and Corti (2005), 223-243.

Marius Victorinus (1888) De definitionibus , ed. Th. Stangl, Tulliana et Mario-Victoriniana,
Munich.

Marmo, C. (ed) (1997) Vestigia, imagines, verba. Semiotics and Logic in Medieval Theological Texts
(XTIth-XIVth Century), Turnhout (Semiotic and Cognitive Studies 4).

- (2003) ‘Pascasio Radberto, i grammatici e la semiotica eucaristica in Ovidio Capitani.

Quaranta anniper la storia medievale, ed. C. De Matteis, Bologna, II, 257-69.

- (2005) ‘II “simbolismo” altomedievale: Tra controversie eucaristiche e conflitti di potere’

in Comunicare e significare nellAlto Medioevo, Spoleto, 765-819 (Proceedings of the 52nd
Settimana di Studio del Centro Italiano di Studi suH’Alto Medioevo).

Martin of Dacia (1961) Opera, ed. H. Roos, Copenhagen (Corpus philosophorum danicorum
medii aevi 2).

Martin, C. J. (1991) ‘The Logic of Negation in Boethius’, Phronesis 36, 277-301.

- (2004) ‘An Amputee Is Able to Walk: The Role of the Categories in the Development of

Abaelard’s Theory of Possibility’ in Biard and Rosier-Catach (2004), 225-242.

Maximus the Confessor (1988) Maximi Confessoris Ambigua ad Iohannem iuxta Iohannis Scotti
Eriugenae Latinam interpretationem, ed. E. Jeauneau, Turnhout (CCSG 18).

Meier-Oeser, S. (1997) Die Spur des Zeichens: Das Zeichen und seine Funktion in der Philosophic
des Mittelalters und der friihen Neuzeit, Berlin and New York.

- (2004) ‘Sprache und Bilder im Geist: Skizzen zu einem philosophischen Langzeitprojekt’,

Philosophisches Jahrbuch 111. Jahrgang, 2. Halbband, 312-342.

Mews, C. (1987) ‘Aspects of the Evolution of Peter Abelard’s Thought on Signification and Predi¬
cation’, in Gilbert de Poitiers et ses contemporains, ed. A. de Libera and J. Jolivet, 15-41.

- (1992) ‘Nominalism and Theology before Abaelard: New Light on Roscelin of Com-

piegne’, Vivarium 30, 4-33.

- (2002) ‘Bruno of Rheims and Roscelin of Compiegne on the Psalms’ in M. W. Herren et

al. (eds) Latin Culture in the Eleventh Century, Turnhout, 129-152.

- (2005) ‘Logica in the Service of Philosophy: William of Champeaux and His Influence’ in

R. Berndt (ed) Schrift, Schreiber, Schenker. Studien zur Abtei Sankt Viktor zu Paris und zu den
Viktorinem, Berlin (Corpus victorinum. Instrumenta 1, 61-101).

Meyer, B. (1993) ‘Synecdoque et tradition’, Histoire epistemologie langage 15(11), 7-37.

Michael of Marbasio (1995) Summa de modis significandi, ed. L. G. Kelly, Stuttgart and Bad
Cannstatt (Grammatica speculativa 5).

Moran, D. (1989), The Phibsophy of John Scottus Eriugena. A Study of Idealism in the Middle Ages.
Cambridge.

Muckle, J. T. (1933), ‘Isaac Israeli’s Definition of Truth’, Archives dhistoire doctrinale et litteraire
du moyen Age 8, 5-8.

Murdoch, J. (1979) ‘Propositional Analysis in Natural Philosophy’, Synthese 40, 117-146.

Bibliography / Vivarium 45 (2007) 374-392

387

Mure, D. (1998) ‘Anonymus Pragensis on Equivocation, CIMAGL 68, 63-97.

Mutschmann, H. (1891) ‘Zur Ubersetzertatigkeit des Nicolaus von Regnium (zu Paris lat
14700)’, Berliner Philologische Wochenschrift 22, 691—692.

Narbonne, J.-M. and L. Langlois (eds) (1999) La metaphysique. Son histoire, sa critique, ses enjeux,
Paris and Quebec.

Nicholas of Autrecourt (2001) Correspondance. Articles condamnes, trans. C. Grellard, Paris.

Nightingale, A. (2002) ‘Distant Views: “Realistic” and “Fantastic” Mimesis in Plato’ in J. Annas
and C. Rowe (eds) New Perspectives on Plato, Modern and Ancient, Cambridge, Mass, and
London, 227-247.

Normore, C. (1987) ‘The Tradition of Mediaeval Nominalism’ in J. F. Wippel (ed) Studies in
Medieval Philosophy, Washington, D.C., 201-217.

Norris Clarke, W. (1982) ‘The Problem of the Reality and Multiplicity of Divine Ideas’ in D. J.
O’Meara (ed) Neoplatonism and Christian Thought, Albany, N.Y., 109-127.

Nuchelmans, G. (1973) Theories of the Proposition. Ancient and Medieval Conceptions of the
Bearers of Truth and Falsity, Amsterdam and London (North-Holland Linguistic Series 8).

- (1988) ‘The Distinction actus exercitus / actus significatus in Medieval Semantics’ in

Kretzmann (1988), 37-90.

O’Donnell, J. R. (1939) ‘Nicholas of Autrecourt’, Mediaeval Studies 1 , 179-280 [edition of
Exigit ordo].

Oeing-Hanhoff, L. (1982) ‘Res comme concept transcendental et sur-transcendental’ in Fattori
and Bianchi (eds) (1982), 285-296.

O’Meara, J. (ed) (1987) John Scotus Eriugena: Periphyseon (the Division of Nature), trans. I. P.
Sheldon-Williams, Montreal and Washington.

Osbern of Gloucester (1996) Derivationes, ed. F. Bertini et al. Spoleto.

Panaccio, C. (1992) ‘From Mental Word to Mental Language’, Philosophical Topics 20, 125-147.

- (1997) ‘Angel’s Talk, Mental Language, and the Transparency of the Mind’ in C. Marmo

(ed) Vestigia, imagines, verba: Semiotics and Logic in Medieval Theological Texts (XTIth-XTVth Cen¬
tury), Turnhout, 323-336.

- (1999) Le discours interieur. De Platon a Guillaume d’Ockham, Paris.

- (2001) ‘Reponses de l’Auteur. De quelques variations sur un theme seculaire’, Laval

theologique etphilosophique 57, 261-276.

- (2003) ‘Debates on Mental Language in the Early Fourteenth Century’ in H. A. G. Braak-

huis and C. H. Kneepkens (eds) Aristotle’s Peri Hermeneias in the Latin Middle Ages, Gronin¬
gen and Haren, 85-101.

Paul of Venice (1976) Expositio in libros posteriorum Aristotelis. Venice 1477, repr. New York.

- (1979) Logica Magna. Part I, Fascicule I, ed. N. Kretzmann, Oxford (Classical and Medi¬
eval Logical Texts 2).

Pepin, J. (1973) ‘Mysteria et Symbola dans le commentaire de Jean Scot sur 1’evangile de Saint
Jean’ in J. O’Meara and L. Bieler (eds) The Mind of Eriugena, Dublin, 16-30.

- (1976a) Aspects theoriques du symbolisme dans la tradition dionysienne. Antecedents et

nouveautes’ in Simboli e simbologia nell’Alto Medioevo, I, Spoleto, 33-79 (Proceedings of the
23rd Settimana di Studio del Centro Italiano di Studi sull’Alto Medioevo).

- (1976b) Saint Augustin et la dialectique, Villanova.

Perler, D. (2002), Theorien der Intentionalitat im Mittelalter, Frankfurt.

Peter Abelard (1836) Ouvrages inedits d’Abelard, ed. V. Cousin, Paris.

- (1919-1933) Peter Abaelardsphilosophische Schriften, Munster (BGPTMA 21).

388

Bibliography / Vivarium 45 (2007) 374-392

- (1927) Glossae super Peri ermenias in B. Geyer (ed) Peter Abaelardsphilosophische Schriften,

Munster.

- (1954) Scritti filosofici , ed. M. Dal Pra, Rome and Milan.

- (1969a) Opera theologica I and II, ed. E. M. Buytaert, Turnhout (CCCM 11, 12).

- (1969b) Scritti di logica , ed. M. Dal Pra, Florence (= rev. ed. of Peter Abelard 1954).

- (1970) Dialectica , ed. L. M. de Rijk, Assen (2nd ed, rev.).

Peter Helias (1993) Summa super Priscianum, ed. L. Reilly, Toronto (Studies and Texts 113).

Peter of Spain (1972) Tractatus called afterwards Summule Logicales , ed. L. M. de Rijk, Assen.

- (1992) Syncategoreumata , ed. L. M. de Rijk, trans. J. Spruyt, Leiden, New York and Cologne.

Philoponus (1898) Philoponi (olim Ammonii) In Aristotelis Categorias Commentarium , ed.
A. Busse, Berlin (CAG XIII, 1).

Pickave, M. (ed) (2003) Die Logik des Transzendentalen , Berlin and New York.

Pinborg, J. (1962) ‘Das Sprachdenken der Stoa und Augustins Dialektik’, Classica et mediaevalia
23, 148-177.

- (1967) Die Entwicklung der Sprachtheorie im Mittelalter , Munster (BGPTMA 42).

Pinchard, B. (1987) Metaphysique et semantique. Autour de Cajetan. Etude [texte] et traduction du
“De Nominum Analogia ”, Paris.

Pini, G. (2003) The Transcendental of Logic: Thirteenth-Century Discussions on the Subject Matter
of Aristotle’s “Categories” in Pickave (ed) (2003), 140-159.

Porphyry (1887) Isagoge et in Aristotelis Categorias commentarium , ed. A. Busse, Berlin (CAG 4.1).

- (1887) Porphyrii In Aristotelis Categorias Expositio per interrogationem et responsionem , ed.

A. Busse, Berlin (CAG IV, 1).

- (1966) Isagoge , translatio Boethii, ed. L. Minio-Paluello, Bruges and Paris (AL I, 6-7).

Porro, P. (1994) ‘II Sextus latinus e l’immagine dello scetticisme antico nel medioevo’, Elenchos 2,
229-253.

Pouillon, D. H. (1939) ‘Le premier traite des proprietes transcendentales’, Revue neoscholastique
de philosophic 42, 40-77.

Priest, G. and S. Read (1981) ‘Ockham’s Rejection of Ampliation’, Mind 90, 274-279.

Priscian, (1855-1859) Prisciani Caesarensis Institutionum Grammaticarum libri XVII, 2 vols.
[vol. 1 Books I—XII; vol. 2 Books XIII—XVIII], ed. J. Hertz, Leipzig (GL, II, III).

Proclus (1899, 1901) In Platonis Rem publicam Commentarii , ed. W. Kroll, Leipzig.

- (1903-1906) In Platonis Timaeum Commentaria , ed. E. Diehl, Leipzig.

- (1908) In Platonis Cratylum Commentaria , ed. G. Pasquali, Leipzig.

- (1987) Proclus Commentary on Plato’s Parmenides, trans. G. R. Morrow and J. M. Dillon,

Princeton, N.J.

Pseudo-Dionysius (1857) Opera (PG?j).

Purcell, W. M. (1987) ‘ Transsumptio : A Rhetorical Doctrine of the Thirteenth Century’, Rhetorica
5,369-410.

Quintilian (1970) M. Fabii Quintiliani Institutionis Oratoriae libri duodecim , 2 vols, ed. M. Win-
terbottom, Oxford.

- (1975) Institutio oratoria , ed. J. Cousin, Paris.

Radulphus Brito (1980) Quaestiones super Priscianum Minorem, ed. H. W. Enders and J. Pin¬
borg, Stuttgart and Bad Cannstatt (Grammatica speculativa 3.1-2).

Read, S. (ed) (1993), Sophisms in Medieval Logic and Grammar , Dordrecht, Boston and London.

Remigius of Auxerre (1912) Commentum Einsidlense in Donati artem minorem, recensio brevis,
ed. W. Fox, Leipzig.

Bibliography / Vivarium 45 (2007) 374-392

389

Rist, J. M. (1994), Augustine. Ancient Thought Baptized, Cambridge.

Robert Grosseteste (1981) Commentarius in Posteriorum Analyticorum libros , ed. R Rossi,
Florence.

Robert Holcot (1990) ‘ Quodlibeta 1.6’, ed. Dominik Perler, in Satztheorien. Texte zur Sprach-
philosophie undWissenschaftstheorie im 14. Jahrhundert, Darmstadt (Texte zur Forschung 57),
226-249.

Roger Bacon (1988) Roger Bacon. Compendium of the Study of Theology, ed. and trans. T. S.
Maloney, Leiden.

Robert Kilwardby (1976) De ortu scientiarum, ed. A. G. Judy, London (Auctores britannici
medii aevi 4).

- (1982) Quaestiones in librum tertium Sententiarum: Christologie, ed. E. Gosser, Munich

(Veroffendichungen der Kommission fur die Herausgabe ungedruckter Texte aus der mittelal-
terlichen Geisteswelt).

- (1986) Quaestiones in librumprimum Sententiarum, ed. J. Schneider, Munich (Veroffendichungen

der Kommission fur die Herausgabe ungedruckter Texte aus der mittelalterlichen Geisteswelt 13).

- (1992) Quaestiones in librum secundum Sententiarum, ed. G. Leibold, Munich (Veroffendichungen

der Kommission fur die Herausgabe ungedruckter Texte aus der mittelalterlichen Geisteswelt 16).

Roques, R. (1962) ‘Symbolisme et theologie negative chez le pseudo-Denys’ in Structures
theologiques de la gnose a Richard de Saint-Victor. Essais et analyses critiques, Paris, 164-179
(orig. ed. in Bulletin de PAssociation Bude, 1957).

- (1967) ‘Teratologie et theologie chez Jean Scot Erigene’ in Melanges ojferts a M.-D. Chenu,

maitre en theologie , Paris, 1967, 419-437 (repr. in R. Roques, Libres senders vers Verigenisme,
Rome, 1975, 13-43).

Rosemann, P. W. (1998) ‘ Aliquid : Ein vergessenes Transzendentale’ in Aertsen and Speer, 529-
537.

Rosen, S. (1983) Plato’s Sophist. The Drama of Original and Image , New Haven and London.

Rosier-Catach, I. (1995a) ‘Res significata et modus significandi: Les implications d’une distinc¬
tion medievale’ in S. Ebbesen (ed) Sprachtheorien in Spdtandke und Mittelalter, Tubingen,
135-168.

- (1995b) ‘Henri de Gand, le De dialectica d’Augustin et 1’imposition des noms divins’,

Documend e studi sulla tradizione filosofica medievale 6 , 145-153.

- (1997) ‘Prata Rident’ in A. de Libera, A. Elamrani-Jamal and A. Galonnier (eds) Lan-

gages et Philosophie: Hommage a jean Jolivet, Paris, 155-176 (Etudes de philosophic
medievale 74).

- (1999a) ‘La notion de translatio, le principe de compositionalite et l’analyse de la predica¬
tion accidentelle chez Abelard’ in J. Biard (ed) Langage, Sciences , Philosophie au XII e siecle,
Paris, 125-164.

- (1999b) ‘Aristotle and Augustine. Two Models of Occidental Medieval Semantics’ in H. S.

Gill and G. Manetti (eds) Signs and Signification, New Delhi, vol. II, 41-62.

- (2003a) ‘Abelard et les grammairiens: sur la definition du verbe et la notion d’inherence’

in P. Lardet (ed) La tradition vive, Melanges d’histoire des textes en I’honneur de Louis Holtz,
Turnhout, 143-159.

- (2003b) ‘Priscien, Boece, les Glosulae in Priscianum, Abelard: Les enjeux des discussions

autour de la notion de consignification’, Histoire epistemologie langage 25, 55-84.

- (2003c) ‘Abelard et les grammairiens: sur le verbe substantif et la predication’, Vivarium

41, 176-248.

390

Bibliography / Vivarium 45 (2007) 374-392

- (2004a) ‘Les discussions sur le signifie des propositions chez Abelard et ses contemporains’

in Maieru and Valente (2004), 1-34.

- (2004b) ‘The Glosulae in Priscianum and Its Tradition in N. McLelland and A. Linn (eds)

Papers in Memory of Vivien Law, Munster (Henry Sweet Society Studies in the History of
Linguistics), 81-99.

- (forthcoming (a)) ‘Les Glosulae super Priscianum : semantique et universaux’, Documenti e

studi per la storia della filosofia medievale: Gli universali.

- (forthcoming (b)) ‘“Vox” and “Oratio” in Early Twelfth Century Grammar and Dialectic’

in I. Rosier-Catach (ed) Proceedings of the Conference Les Glosulae in Priscianum, William of
Champeaux, Abelard, Arts du langage et theologie aux confins des XILXIT siecles', Turnhout.

Rosier-Catach, I. and S. Ebbesen (2004) ‘Petrus de Alvernia + Boethius de Dacia: Syllogizantem
PONENDUM EST TERMINOS’, CIMAGLl'b , 161-218.

Ruef, H. (1981) Augustin iiber Semiotik und Sprache. Sprachtheoretische Analysen zu Augustins
Schrifi “De Dialectical', Berne.

Saffrey, H. D. (1978) ‘Nouveaux liens objectifs entre le pseudo-Denys et Proclus’, Revue des
sciencesphilosophiques et theologiques 63, 6-13.

Salmond, S. D. F. (1898) John of Damascus. Exposition of the Orthodox Faith [Engl, transl.]
in R. Schaff and H. Wace (eds) Nicene and Post-Nicene Fathers of the Christian Church,
2nd Series, 9, Edinburgh (available from www.monachos.net).

Samstag, K. (1929), Die Dialektik des Johannes Skottus Eriugena. Inaugural dissertation, Munich.

Schmitt, C. B. (1972) Cicero Scepticus. A Study of the Influence of the Academics in the Renaissance,
The Hague.

- (1983) ‘The Rediscovery of Ancient Skepticism in Modern Times’ in M. Burnyeat (ed)

The Skeptical Tradition , Berkeley, Los Angeles and London, 225-231.

Schmitz, H. R. (1981) ‘Un transcendental meconnu’, Cahiers Jacques Maritain 2, 21-51.

Schneider, J. H. J. (1993) ‘Al-Farabis Kommentar zu De interpretatione des Aristoteles: Ein
Beitrag zur Entwicklung der Sprachphilosophie im Mittelalter’ in Craemer-Ruegenberg and
Speer (1993), vol. 2, 687-738.

Schneider, R. (1910) Librorum Apollonii deperditorum fragmenta , Leipzig (GG 11.3).

Schum,W. (1887) Beschreibendes Verzeichnis der Amplonianischen Handschriften zu Erfurt , Berlin.

Searle, J. (1969) Speech Acts: An Essay in the Philosophy of Language, Cambridge.

Sedley, D. (2003) Plato's Cratylus, Cambridge.

Sedulius Scotus (1977) In Donati artem maiorem (II 2), ed. B. Lofstedt, Turnhout (CCCM 40B).

- (1977) In Priscianum , ed. B. Lofstedt, Turnhout (CCCM 40C).

Sheppard, A. D. R. (1980) Studies on the 5th and 6th Essays of Proclus' Commentary on the Repub¬
lic, Gottingen.

Shimizu, T. (1995) ‘From Vocalism to Nominalism: Progression in Abaelard’s Theory of
Signification’, Didascalia 1, 15-46.

Simon of Faversham (1984) Quaestiones super libro Elenchorum, ed. S. Ebbesen, T. Izbicki,
J. Longeway, F. del Punta, E. Serene and E. Stump, Toronto.

Simplicius (1971) Commentaire sur les Categories d'Aristote. Traduction de Guillaume de Moerbeke,
2 vols, ed. A. Pattin, Louvain and Paris.

Sirridge, M. (1988) ‘Robert Kilwardby as “Scientific Grammarian”’, Histoire epistemologie,
langage X (I), 1-21.

- (1990) ‘Robert Kilwardby: Figurative Constructions and the Limits of Grammar’ in

S. Ebbesen, G. L. Bursill-Hall and K. Koerner (eds) De Ortu Grammaticae, Amsterdam and
Philadelphia, 321-337 (Studies in the History of the Language Sciences 43).

Bibliography / Vivarium 45 (2007) 374-392

391

- (1995) ‘The Science of Language and Linguistic Knowledge’ in S. Ebbesen (ed) Sprach-

theorien in Spdtantike und Mittelalter, Tubingen (Geschichte der Sprachtheorie 3).

- (1999) ‘ Quam videndo intus dicimus : Seeing and Saying in De TrinitateXV ’ in S. Ebbesen

and R. L. Friedman (eds) Medieval Analyses in Language and Cognition, Copenhagen (Histo-
risk-filosofiske meddelelser 77), 317-330.

Sorabji, R. (1990) Aristotle Transformed, London.

Spade, R V. (1980a) ‘Synonymy and Equivocation in Ockham’s Mental Language ’,Journal of the
History of Philosophy 18, 9-22.

- (1980b) ‘Review of: M. Tweedale, Abailard on Universals’, Nous 14, 479-483.

Speca, A. (2001) Hypothetical Syllogistic and Stoic Logic, Leiden and Boston (Philosophia antiqua 87).

Spruyt, J. (2003) ‘Twelfth Century Glosses on Aristotle’s Category of “Relational Being” (“Pros
ti”)’ in Biard and Rosier-Catach (2003), 329-345.

Stanley, R. P. (1997) ‘Hipparchus, Plutarch, Schroder and Hough’, American Mathematical
Monthly 104, 344-350.

Stone, M. W. F. (2001) ‘Augustine and Medieval Philosophy’ in E. Stump and N. Kretzmann
(eds) The Cambridge Companion to Augustine, Cambridge.

Strange S. and J. Zupko (eds) (2004) Stoicism: Traditions and Transformations, Cambridge.

Synan, E. A. (1967) ‘Master Peter Bradlay on the Categories', Mediaeval Studies 29, 273-327.

Tachau, K. (1988) Vision and Certitude in the Age of Ockham, Leiden.

Tatuinus (1968) Arsgrammatica, ed. M. de Marco, Turnhout (CCSL 133).

Thierry of Chartres (1971) Commentaries on Boethius by Thierry of Chartres and His School, ed.
N. M. Haring, Toronto (Studies and Texts 20).

Thomas Aquinas (1970) Quaestiones disputatae de veritate, ed. A. Dondaine, Rome (Opera
omnia XXII).

- (1976) De fallaciis in Opera omnia 43, Rome.

Thomas Claxton (1943) Lhomae de Claxton O.P. (CA. 1400): Quaestiones de distinctione inter esse
et essentiam reali atque de analogia ends, ed. M. Grabmann, Acta Pontificiae Academiae Roma-
nae 5. Thomae Aquinatis et Religionis Catholicae 8 (New Series), 92-153.

Thomas of Erfurt (1972) Grammatica Speculativa, ed. G. L. Bursill-Hall, London.

Thomas, R. (ed) (1980) Peter Abaelard, Person, Werk und Wirkung, Trier.

Trouillard, J. (1981) ‘Le symbolisme chez Proclos’, Dialogues d'histoire ancienne 7, 297-308.

Valente, L. (2005) “‘Ilia quae transcendunt generalissima”: Elementi per una storia latina dei
termini trascendentali (XII secolo)’ in P. Porro (ed) Metaphysica — sapientia—scientia divina:
soggetto e statuto della filosofia prima nel Medioevo [= Quaestio 5 (2005), 217-239].

- (2006) “‘Ens, unum, bonum”: elementi per una storia dei trascendentali in Boezio e nella

tradizione boeziana del XII secolo’ in S. Caroti, R. Imbach, Z. Kaluza, G. Stabile and L. Stur-
lese (eds) Ad Ingenii Acuitionem. Studies in Honour of Alfonso Maierii, Louvain-la-Neuve
(FIDEM, Textes et etudes du moyen age, 38), 483-545.

- (in press) Logique et theologie. Les ecolesparisiennes entre 1150 et 1220, Paris.

Ventimiglia, G. (1997) Dijferenza e contraddizione. IIproblema dell'essere in Tommaso dAquino:
esse, diversum, contradictio, Milan (Metafisica e storia della metafisica 17).

- (1998) ‘Die Transzendentalienlehre des Thomas von Aquin: Denktraditionen, Quellen,

Eigenheiten’ in J. A. Aersten and A. Speer (eds) (1998), 522-528.

Vicaire M. H. (1937) ‘Les Porretains et 1’Avicennisme avant 1215’, Revue des sciences phi-
losophiques et theologiques 26, 449-482.

Villela-Petit, M. (1991) ‘La question de l’image artistique dans le Sophiste' in P. Aubenque (ed)
Etudes sur le Sophiste de Platon, Naples, 53-90.

392

Bibliography / Vivarium 45 (2007) 374-392

Vincent Ferrer (1977) Tractatus de suppositionibus, edj. A.Trentman, Stuttgart and Bad Cannstatt.

Von Perger, M. (2005), ‘Eriugenas Adaptation der aristotelischen Kategorienlehre’ in D. Perler
and U. Rudolph (eds) Logik und Theologie. Das Organon im arabischen und im lateinischen
Mittelalter , Leiden and Boston, 239-303.

Walter Burley (1497a) Liber Praedicamentorum, in Super Artem Veterem Porphyrii et Aristotelis,
Venice (repr. Frankfurt, 1967).

- (1497b) Liber Perihermeneias , in Super Artem Veterem Porphyrii et Aristotelis, Venice

(repr. Frankfurt, 1967).

- (Burleigh) (1955) De Puritate Artis Logicae Tractatus Longior with a Revised Edition of the

Tractatus Brevior, ed. P. Boehner, St. Bonaventure, N.Y., Louvain and Paderborn.

- (1973) Commentarius in librum Perihermeneias, ed. S. F. Brown in ‘Walter Burleys Middle

Commentary on Aristotle’s Perihermeneias, Franciscan Studies 33, 45-134.

- (1974) Quaestiones super librum Perihermeneias , ed. S. F. Brown, Franciscan Studies 34,

200-295.

- (1998) Tractatus de universalibus , ed. H.-U. Wohler, Stuttgart and Leipzig.

- (2003) Super tractatum fallaciarum , ed. S. Ebbesen, Burley on Equivocation in His Com¬
panion to a ‘ Tractatus Fallaciarum and in His Questions on the Flenchi\ CIMAGL 74, 151-207.

Weber, S. (2003) Richard Billingham, De Consequents mit Toledo Kommentar, Amsterdam.

Wiesner, J. (ed) (1987) Aristoteles, Werk und Wirkung, Berlin and New York.

William Crathorn (1988) Crathorn. Qudstionen zu ersten Sentenzenbuch. Einfuhrung und Text,
ed. F. Hoffmann, Munster (BGPTMA Neue Folge Bd. 29).

- (1999) Glosae super Boethium, ed. L. Nauta, Turnhout (CCCM 145).

William of Ockham (1974) Summa logicae, ed. P. Boehner, G. Gal and S. Brown, St. Bonaven¬
ture, N.Y. (Opera philosophica I).

- (1979) Expositio super libros Elenchorum, ed. F. del Punta, St. Bonaventure, N.Y. (Opera

philosophica 3).

William of Sherwood (1995) Introductiones in Logicam: Einfuhrung in die Logik , ed. and trans.
H. Brands and C. Kann, Hamburg.

Wilson, C. (1956) William Heytesbury: Medieval Logic and the Rise of Mathematical Physics,
Madison.

Wittwer, R. (2002) ‘Zur Lateinischen Uberlieferung von Sextus Empiricus’ FIYPP^NEIOI-
YnOTYnftZEII’, Rheinisches Museum fur Philologie, n.f. 145, 366-373.

Wolfson, H. A. (1973) ‘The Amphibolous Terms in Aristotle, Arabic Philosophy and Maimo-
nides’ in his Studies in the History of Philosophy and Religion, vol. 1, ed. I. Twersky and G. H.
Williams, Cambridge, Mass, and London, 455-477.

Zambon, M. (2002), Porphyre et le moyen-platonisme, Paris.

- (2003) ‘Aristotelis Platonisque sententias in unam revocare concordiam. II progetto filosofico

boeziano e le sue fonti’, Medioevo 28, 17-49.

Zupko, J. (1993) ‘Buridan and Skepticism’, Journal of the History of Philosophy 31 (2), 191-221.

- (2003) John Buridan: Portrait of a Fourteenth-Century Arts Master, Notre Dame, Ind.

JOHN VC YU 11

BRILL

A Companion to John Wyclif
Late Medieval Theologian

Edited by Ian C. Levy

• September 2006

• ISBN 978 9004150 07 2

»Hardback (xxii, 490 pp.)

• List price EUR 95.- / US$ 124.-
Brill’s Companions to the Christian
Tradition, 4

B The Companion to John Wyclif contains
eight substantial essays covering the
central aspects of John Wyclif’s life
and thought. The volume’s authors
have drawn on an extensive amount of
primary material, as well as the most
recent secondary sources, so as to present a comprehensive picture of
Wyclif in his times. Topics covered include a detailed life and career
of Wyclif, and close analyses of his logic and metaphysics; doctrine of
the Trinity and Christology; political views; Christian life and piety;
sacraments; the Bible; and an examination of his medieval opponents.
Experts and students alike will profit from these in-depth studies all
of which provide a view of Wyclif in his late medieval context. For
those not already familiar with Wyclif this volume will serve as an
excellent introduction; and those with greater expertise will find fresh
appraisals which may, in turn, lead to further research.

Ian Christopher Levy (Marquette, 1997) is Assistant Professor of
Theology at Lexington Theological Seminary. He has published articles
on various aspects of John Wyclif’s thought. He has also published
an abridged translation of Wyclif’s On the Truth of Holy Scripture
(2001); and a monograph on Wyclif’s eucharistic theology: John Wyclif:
Scriptural Logic, Real Presence and the Parameters of Orthodoxy (2003). He
is currently completing a volume of medieval commentaries on the
Epistle to the Galatians.

For our Conditions of Sales please see www.brill.nl

To order and for more information e-mail brill@turpin-distribution.com
or cs@brillusa.com (for customers in North America).

f 1 s *

www.brillonline.nl

Brill Online -

Brill’s Online Reference Works Platform
Brill Online will bring together major online
reference works from Brill under one platform

As an academic publisher with a long
and proud history serving the academic
community around the world, Brill is
delighted to announce the launch of our new
online reference works platform, Brill Online.
Brill’s reference works have long enjoyed a
reputation for quality, with a number of our
major reference works selected as CHOICE
Outstanding Academic Titles.

n* c /*>

' 68 ^

THE FOLLOWING TITLES ARE ALREADY AVAILABLE

- Africa Yearbook

- The Brill Dictionary of Religion

- Encyclopaedia of Islam

- Encyclopaedia of the Qur’an

- Encyclopaedia of Judaism, 2nd edition

- Encyclopedia of Women in Islamic Cultures
-Jacoby Online

- Pauly Online

Online database accessible via Brill Online:

- World Christian Database

BENEFITS OF WWW.BRILLONLINE.NL INCLUDE

- Fully Unicode compliant, to facilitate the display of foreign languages

- Ability to cross-search multiple titles

- COUNTER-compliant usage statistics

- ATHENS authentication

How to Order

For more information about Brill’s
Online Reference works, consortia and
other pricing options, please send an
e-mail to brillonline@brill.nl

BRILL

Online subscriptions may start at any time during the year. A 30 day free trial is
available to all institutions. Please contact brillonline@brill.nl

Theological Quodlibeta in the Middle
Ages: The Thirteenth Century

Edited by Christopher Schabel

•July 2006

• ISBN 978 9004123 33 5

• Hardback (xiv, 564 pp.)

• List price EUR 125.- / US$ 163.-

• Brill’s Companions to the Christian
Tradition, 1

This is the first of two volumes on
theological quodlibeta, records of special
disputations held before Christmas
and Easter ca. 1230-1330, mostly at the
University of Paris, in which audience
members asked the great masters of
theology the questions for debate, questions de quolibet, “about
anything.” The variety of the material and the authors’ stature make
the genre uniquely fascinating.

In Volume I, chapters by acknowledged experts introduce the genre,
cover the quodlibeta of Thomas Aquinas, Henry of Ghent, Giles of
Rome, Godfrey of Fontaines, and 13th-century Franciscans, and
demonstrate how the masters used quodlibeta to construct and
express their authority on issues from politics and economics to two-
headed monsters.

For all those interested in medieval studies, especially intellectual
history.

Chris Schabel, Ph.D. (1994), Iowa, teaches at the University of
Cyprus. Specializing in intellectual history and the Latin East, he
recently published Theology at Pahs 1316-1345 (2000), The Synodicum
Nicosiense (2001), and, with A. Nicolaou-Konnari, Cyprus - Society and
Culture 1191-1374 (Brill, 2005).

For our Conditions of Sales please see www.brill.nl

To order and for more information e-mail brill@turpin-distribution.com
or cs@brillusa.com (for customers in North America).

ONLINE JOURNALS

C / 0

; 68i

BRILL

Electronic Access to your Journal
at no extra charge

Brill is pleased to announce that you - as an institutional
subscriber to the printed version of this journal - can access the
electronic version at no extra cost.

How to gain access to Brills online journals

Institutions

You can go to the URL as * r

given in this issue or to j »»»

www.brill.nl and search/ ~ -

browse for the title of your HrT” --- * l

journal in the search bar. “T i.

Before you can access Eim : 8

online journals you
need to register with

Ingenta first (at http:// --<

www.ingentaconnect.

com/register/institutional ). Please click ‘register’ and fill out
the registration form. You will need to fill out your IP-address
range(s) or your User-
ID and Password. After
completion of this
form you will get a CID
registration number
and you will see your
IP address range(s).

Now you can activate
your subscription(s) by
logging in to activate
institutional and fill out
the institutional activation form, including your CU-prefixed
Customer Number. This number can be found on the address
label, which is sent with your issue(s), on the (renewal) invoice
(listed as “ship-to”), or you may ask your subscription agent to
supply you with your Brill customer number. Please click ‘enable
access’ to complete the process. On your screen you will see a list
of the journals you have electronic access to.

Institutional Registratior

Our Journals are hosted by Ingenta and are available via a number of
journal agents

To order and for more information e-mail brill@turpin-distribution.com
or cs@brillusa.com (for customers in North America).

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "VIVARIUM"

See other formats